System Catalog - Ocient Documentation

The system catalog exposes read-only virtual tables that contain metadata about the system.

Alphabetical List of Tables

sys.active_operator_instances
sys.active_query_scheduler_info
sys.addendum_directories
sys.all_operator_instances
sys.association_rules_models
sys.average_bb_sizes
sys.average_column_sizes
sys.bagging_models
sys.boosting_models
sys.channel_endpoint_parameters
sys.clusters
sys.column_cardinalities
sys.column_distributions
sys.columns
sys.columns_compression_info
sys.completed_operator_instances
sys.completed_queries
sys.compute_configurations
sys.config
sys.connectivity_pool_participants
sys.connectivity_pools
sys.constraints
sys.current_osn
sys.data_usage
sys.databases
sys.dbscan_models
sys.decision_tree_models
sys.degraded_segment_groups
sys.feature_selection_models
sys.feature_selection_steps
sys.feedforward_network_models
sys.function_parameters
sys.function_signatures
sys.functions
sys.gaussian_discriminant_analysis_models
sys.gaussian_mixture_models
sys.global_map_table_info
sys.gradient_boosted_trees_models
sys.group_roles
sys.groups
sys.index_advisor_column_usages
sys.index_advisor_enabled
sys.index_advisor_index_usages
sys.index_columns
sys.index_recommendations
sys.index_usage
sys.indexes
sys.k_means_models
sys.k_nearest_neighbors_models
sys.linear_combination_regression_models
sys.linear_discriminant_analysis_models
sys.load_errors
sys.load_events
sys.load_metrics
sys.locations
sys.locks
sys.logistic_regression_models
sys.lts_cluster_info
sys.machine_learning_model_options
sys.machine_learning_models
sys.mcp_tools
sys.merge_eligibilities
sys.merge_policies
sys.metric_levels
sys.mlp_models
sys.multiple_linear_regression_models
sys.naive_bayes_models
sys.network_interface_models
sys.network_interface_usage_types
sys.node_clusters
sys.node_config
sys.node_network_interfaces
sys.node_status
sys.nodes
sys.nonlinear_regression_models
sys.oidc_integrations
sys.oidc_sessions
sys.op_inst_debug_info
sys.orphaned_segments
sys.osn_acquisitions
sys.pipeline_current_throughput
sys.pipeline_errors
sys.pipeline_errors_historical
sys.pipeline_events
sys.pipeline_events_historical
sys.pipeline_files
sys.pipeline_files_historical
sys.pipeline_files_pending
sys.pipeline_functions
sys.pipeline_metrics
sys.pipeline_metrics_current
sys.pipeline_metrics_historical
sys.pipeline_metrics_info
sys.pipeline_partitions
sys.pipeline_tables
sys.pipeline_tasks
sys.pipelines
sys.plans
sys.polynomial_regression_models
sys.principal_component_analysis_models
sys.privileges
sys.procedure_parameters
sys.procedures
sys.queries
sys.random_forest_models
sys.regression_tree_models
sys.reserved_words
sys.result_cache
sys.retention_policies
sys.rights
sys.roles
sys.sample_segment_io_pipelines
sys.scheduled_tasks
sys.schemas
sys.security_integrations
sys.security_settings
sys.segment_batch_events
sys.segment_directories
sys.segment_generation_events
sys.segment_group_transfers
sys.segment_groups
sys.segment_part_inventory
sys.segment_part_redundancy_info
sys.segment_parts
sys.segments
sys.segments_compression_info
sys.service_classes
sys.service_classes_for_user
sys.service_role_channel_endpoints
sys.service_role_status
sys.service_roles
sys.sessions
sys.simple_linear_regression_models
sys.sql_messages
sys.stacking_models
sys.stats_files
sys.stop_osn_reap_requests
sys.storage_capacity
sys.storage_device_files
sys.storage_device_metrics
sys.storage_device_status
sys.storage_scopes
sys.storage_spaces
sys.storage_used
sys.stored_leaf_segment_part_inventory
sys.stored_segments
sys.subtasks
sys.support_vector_machine_models
sys.survival_regression_models
sys.system_information
sys.system_table_columns
sys.system_table_constraints
sys.system_tables
sys.table_cardinalities
sys.tables
sys.tasks
sys.tkt_table_clustering_column_indices
sys.tkt_table_info
sys.undersampling_models
sys.unhealthy_segments
sys.user_groups
sys.user_mappings
sys.user_roles
sys.users
sys.vector_autoregression_models
sys.view_columns
sys.views
sys.vl_columns_compression_info

Catalog Table Details

Configuration

sys.config

This table contains all configuration overrides applied on the system.

Column Name	Column Type	Column Description
scope_type	CHAR	Type of the scope.
scope_id	UUID	Universally Unique IDentifier (UUID) of the storage scope where this override applies.
key	CHAR	String that represents the configuration parameter.
value	CHAR	Value for the configuration parameter key.

sys.node_config

This table contains the effective node configuration for all nodes in the system.

Column Name	Column Type	Column Description
node_id	UUID	Universally Unique IDentifier (UUID) of the node where this configuration exists (sys.nodes).
key	CHAR	String that represents the configuration parameter.
value	CHAR	Value for the configuration parameter key.
restart_required	BOOLEAN	Whether the parameter requires a node restart to take effect.

Connectivity Pools

sys.connectivity_pool_participants

This table contains the connectivity pool participants and their information.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the connectivity pool that contains the participant.
node_id	UUID	UUID of the participant node.
created_at	TIMESTAMP	Timestamp that indicates the creation of the connectivity pool participant.
updated_at	TIMESTAMP	Timestamp that indicates the last update of the connectivity pool participant.
listen_address	CHAR	Address where the participant node listens.
listen_port	INT	Port number where the participant node listens.
advertised_address	CHAR	Address that the participant node advertises to the client.
advertised_port	INT	Optional port number that the participant node advertises to the client.
openapi_port	INT	Optional port number where the API, written according to the OpenAPI Specification (OAS), listens.
openapi_advertised_address	CHAR	Optional address that the OpenAPI participant advertises for callbacks.
openapi_advertised_port	INT	Optional port number that the OpenAPI participant advertises for callbacks.
ssl_certificate_file	CHAR	Optional base name of the SSL certificate file for the participant.
openapi_ssl_certificate_file	CHAR	Optional base name of the SSL certificate file for the OpenAPI endpoint.

sys.connectivity_pools

This table contains the connectivity pools and their information (excluding participants).

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the connectivity pool.
name	CHAR	Name of the connectivity pool.
created_at	TIMESTAMP	Timestamp that indicates the creation of the connectivity pool.
updated_at	TIMESTAMP	Timestamp that indicates the last update of the connectivity pool.
source_address	CHAR	Source address in CIDR notation for the connectivity pool.
source_port	INT	Optional source port of the connectivity pool.
priority	INT	Priority of the connectivity pool.
sso_integration_name	CHAR	Name of the default SSO security integration for the connectivity pool.
require_sso	BOOLEAN	Whether this connectivity pool requires SSO authentication.

Databases

sys.columns

This table contains all columns in each table in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the column.
name	CHAR	Name of the column.
data_type	CHAR	Data type of the column (INT, CHAR, BOOLEAN, etc.).
position	INT	Position of the column in its table.
table_id	UUID	UUID of the table that contains this column (sys.tables).
nullable	BOOLEAN	Specifies whether the values in this column can be NULL.
default_expression	CHAR	Default value of this column when the value is unspecified.
description	CHAR	Description of the column.
type_hint	CHAR	Type hint for the expected type of the column.
ordinal	LONG	Ordinal of the column with respect to its table.
gdc_type	CHAR	Type of GDC applied to this column.
system	BOOLEAN	Specifies whether the column is a system column.

sys.constraints

This table contains the constraints for the database.

Column Name	Column Type	Column Description
id	UUID	The Universally Unique IDentifier (UUID) of the constraint.
table_id	UUID	The UUID of the table (sys.tables).
name	CHAR	The name of the constraint that has to be unique within an individual table.
constraint_type	CHAR	The type of the constraint (PRIMARY_KEY, UNIQUE, FOREIGN_KEY).
column_names	ARRAY(CHAR)	The names of the columns where the constraint applies.
referenced_constraint_id	UUID	The UUID of the referenced constraint. This value is NULL for non-foreign key constraints.
referenced_table_id	UUID	The UUID of the referenced table. This value is NULL for non-foreign key constraints.
referenced_column_names	ARRAY(CHAR)	The names of the columns where the referenced constraint applies. This value is NULL for non-foreign key constraints.

sys.databases

This table contains all databases defined in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the database.
name	CHAR	Name of the database.
created_at	TIMESTAMP	Timestamp that specifies when the database was created.
user_can_view_all_queries	BOOLEAN	Current user can view all queries made in this database

sys.function_parameters

This table contains information for the parameters of user-defined functions in the system.

Column Name	Column Type	Column Description
function_id	UUID	Universally Unique IDentifier (UUID) of the user-defined function (sys.functions).
parameter_name	CHAR	Name of the parameter. This value is NULL if it represents the return value for the function.
ordinal_position	INT	Position of the parameter. This value is 0 if it represents the return value for the function. Otherwise, this value increments from 1 for input parameters.
data_type	CHAR	Data type of the parameter. The default value is NULL.
nullable	BOOLEAN	Whether you can specify this parameter as NULL. If you set this parameter to `true`, then it is nullable.

sys.functions

This table contains all of the user-defined functions in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the user-defined function.
schema	CHAR	Schema of the user-defined function.
name	CHAR	Name of the user-defined function.
sql_expression	CHAR	For SQL user-defined functions, the expression that defines the function.
database_id	UUID	UUID of the database of this user-defined function (sys.databases).
created_at	TIMESTAMP	Timestamp of when this user-defined function was created.
updated_at	TIMESTAMP	Timestamp of when this user-defined function was last updated.
language	CHAR	Language that the user-defined function is written in.
function_type	CHAR	Type of the user-defined function
filename	CHAR	For non-SQL user-defined functions, the name of the file containing the user-defined function code.
function_name_in_external_code	CHAR	For non-SQL user-defined functions, the name of the Java® or Python® function or method that defines the user-defined function.
return_type	CHAR	Return type of the user-defined function.
canonical_name	CHAR	Canonical name of the user-defined function (name_argcount).
deterministic	BOOLEAN	Whether the user-defined function is deterministic (does not execute non-deterministic functions).
description	CHAR	Description of the user-defined function.

sys.global_map_table_info

This table contains the global dictionary compression (GDC) details associated with each compressed column.

Column Name	Column Type	Column Description
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
column_id	UUID	UUID of the column (sys.columns).
compressed_size	INT	Compressed column size (in bytes).
max_count	LONG	Maximum amount of unique column values allowed in the column.
current_count	LONG	Current amount of unique column values in the column.

sys.index_advisor_enabled

This table contains all tables and databases where the index advisor is currently enabled.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the table or database.
name	CHAR	Name of the table or database.
object_type	CHAR	Determines whether this object is a database or table.

sys.index_columns

This table contains the index settings applied to each column.

Column Name	Column Type	Column Description
index_id	UUID	Universally Unique IDentifier (UUID) that represents the index (sys.indexes).
column_id	UUID	UUID that represents the column (sys.columns).
ordinal	INT	Ordinal of the column in the index structure.
tuple_element	CHAR	Specifies the name of the component column that the index applies to if the index exists on a tuple column.
ascending	BOOLEAN	Indicates whether the values in this column are built in ascending order.
column_ordinal	LONG	Ordinal position of this column in the source table.

sys.index_recommendations

This view contains aggregated index recommendations with one recommendation per column that chooses the most frequently suggested index type.

Column Name	Column Type	Column Description
table_name	VARCHAR	The name of the table.
column_name	VARCHAR	The name of the column.
create_index_sql	VARCHAR	The SQL statement to use for index creation.
index_usage_count	BIGINT	The number of times the database uses this index.

sys.index_usage

This view contains index usage information from the persistent index usages table.

Column Name	Column Type	Column Description
table_id	UUID	The Universally Unique IDentifier (UUID) of the table where the index is located.
index_id	UUID	The UUID of the index.
segment_id	UUID	The UUID of the segment with the index.
query_id	UUID	The UUID of the query that could have used the index.
index_used	BOOLEAN	Whether or not the database uses the index.

sys.indexes

This table contains all indexes defined on tables.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the index.
name	CHAR	Name of the index.
index_type	CHAR	Type of the index.
table_id	UUID	UUID of the table for the index (sys.tables).
index_use	CHAR	Intended use for the index by the system.
ngram_size	INT	Size of each N-gram (in bytes).
enabled	BOOLEAN	Whether the index is enabled on this table.

sys.locations

This table contains all of the external table providers.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the location.
name	CHAR	Name of the location.
remote_info	CHAR	Remote information that defines how to talk to the external table provider.
database_id	UUID	UUID of the database of this location (sys.databases).
created_at	TIMESTAMP	Timestamp that represents when this location was created.
updated_at	TIMESTAMP	Timestamp that represents when this location was last updated.

sys.procedure_parameters

This table contains the parameters for SQL stored procedures in the system.

Column Name	Column Type	Column Description
procedure_id	UUID	Universally Unique IDentifier (UUID) of the stored procedure (sys.procedures).
parameter_name	CHAR	Name of the parameter.
ordinal_position	INT	The position of the parameter in the argument list of the stored procedure. The position starts at 1.
data_type	CHAR	Data type of the parameter.
nullable	BOOLEAN	Whether you can specify this parameter as NULL. If you set this parameter to `true`, then it is nullable.

sys.procedures

This table contains all of the stored procedures in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the stored procedure.
schema	CHAR	Schema of the stored procedure.
name	CHAR	Name of the stored procedure.
database_id	UUID	UUID of the database of this stored procedure (sys.databases).
created_at	TIMESTAMP	Timestamp of when this stored procedure was created.
updated_at	TIMESTAMP	Timestamp of when this stored procedure was last updated.
language	CHAR	Language that the stored procedure is written in.
filename	CHAR	For Java and Python stored procedures, the name of the file that the database should load when it needs this stored procedure.
function_name	CHAR	For Java and Python stored procedures, the name of the function that executes the stored procedure.
sql_body	CHAR	For SQL stored procedures, the SQL statements to execute when the database runs this stored procedure.
returns_result	BOOLEAN	For SQL stored procedures, the system sets this value to `true` if the stored procedure returns a result set. The default value is `false`.
canonical_name	CHAR	Name of the stored procedure with the suffix that concatenates the `_` character and the number of parameters.
description	CHAR	User-defined description of the stored procedure.
num_args	INT	Number of arguments the stored procedure accepts.

sys.tables

This table contains all tables defined in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
name	CHAR	Name of the table.
schema	CHAR	Name of the schema.
database_id	UUID	UUID of the database (sys.databases).
storage_space_id	UUID	UUID of the storage space (sys.storage_spaces).
maximum_segment_size_gib	INT	Maximum size of a segment for the table in GiB.
description	CHAR	Detailed description of the table.
streamloader_property_string	CHAR	Additional properties assigned to control the stream loading behavior for the table.
lts_property_string	CHAR	Additional properties assigned to control the Foundation role (formerly “lts role”) behavior for the table.
created_at	TIMESTAMP	Timestamp that represents when the table was created.
altered_at	TIMESTAMP	Timestamp that represents when the table’s structure was last modified via DDL.
rolehostd_version	CHAR	The rolehostd version in which this table was created
software_compatible_version	INT	The rolehostd version this table is compatible with.
creator_id	UUID	The UUID of the user or group who created the table (sys.users).

sys.tkt_table_clustering_column_indices

This table contains the columns defined as part of a clustering index.

Column Name	Column Type	Column Description
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
column_id	UUID	UUID of the column (sys.columns).

sys.tkt_table_info

This table contains the columns that are defined as the time key on each user-defined table and the size of the time bucket in use.

Column Name	Column Type	Column Description
table_id	UUID	The Universally Unique IDentifier (UUID) of the table (sys.tables).
time_column_id	UUID	The UUID of the column that represents the time column in the table (sys.columns).
time_bucket_width	LONG	The width of the time bucket for segmenting data in nanoseconds or days.
time_bucket_width_interval	CHAR	The readable interval for the width of the time bucket for segmenting data.
retention_policy	LONG	The interval for the retention policy for rolling off data in nanoseconds or days.
retention_policy_interval	CHAR	The readable interval of the retention policy for rolling off data.

sys.user_mappings

This table contains user mappings for locations.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the user mapping.
name	CHAR	Name of the user mapping.
local_userid	CHAR	Local user identifier.
remote_userid	CHAR	Remote user identifier.
location_id	UUID	UUID of the location for this user mapping.
database_id	UUID	UUID of the database for this user mapping (sys.databases).
created_at	TIMESTAMP	Timestamp that represents when this user mapping was created.
updated_at	TIMESTAMP	Timestamp that represents when this user mapping was last updated.
remote_password	CHAR	Password for the remote user identifier.

sys.view_columns

This table contains all columns in each view in the system.

Column Name	Column Type	Column Description
id	UUID	The Universally Unique IDentifier (UUID) of the column.
name	CHAR	The name of the column.
data_type	CHAR	The data type of the column.
view_id	UUID	The UUID of the view this column comes from.
nullable	BOOLEAN	Whether or not this column is nullable.
default_expression	CHAR	The default expression of this column, if it exists.
ordinal	LONG	The ordinal of this column.
global_dictionary_compression_table_id	UUID	UUID of the table with the GDC column (sys.tables).

sys.views

This table contains all user-defined database views created on the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the view.
name	CHAR	Name of the view.
database_id	UUID	UUID of the database (sys.databases).
schema	CHAR	Schema name where the view exists.
query	CHAR	Query used to generate the view content.
description	CHAR	Detailed description of the view.
global_dictionary_compression_table_id	UUID	UUID of the table with the GDC column (sys.tables).
created_at	TIMESTAMP	Timestamp that represents the date and time for the creation of the view.
updated_at	TIMESTAMP	Timestamp that represents the date and time for the last update of the view.
creator_id	UUID	The UUID of the user or group who created the view (sys.users/sys.groups).

Index Advisor

sys.index_advisor_column_usages

This table contains column usage data collected by the index advisor for recommending secondary indexes.

Column Name	Column Type	Column Description
database_id	UUID	The Universally Unique IDentifier (UUID) of the database.
database_name	VARCHAR	The name of the database.
table_id	UUID	The UUID of the table.
column_id	UUID	The UUID of the column.
table_name	VARCHAR	The name of the table.
column_name	VARCHAR	The name of the column.
index_type_name	VARCHAR	The recommended index type (e.g., `INDEX_TYPE_INVERTED` for an inverted index).
query_id	UUID	The UUID of the SQL query that triggered this recommendation.
query_time	TIMESTAMP	The timestamp when the query was executed.
ngram_value	INT	The recommended ngram size for NGRAM indexes. This value is NULL for other indexes.
create_index_sql	VARCHAR	The SQL statement to use for index creation.

sys.index_advisor_index_usages

This table contains index usage data collected by the index advisor that tracks whether existing indexes were used.

Column Name	Column Type	Column Description
database_id	UUID	The Universally Unique IDentifier (UUID) of the database.
database_name	VARCHAR	The name of the database.
table_id	UUID	The UUID of the table.
index_id	UUID	The UUID of the index.
segment_id	UUID	The UUID of the segment.
query_id	UUID	The UUID of the SQL query.
query_time	TIMESTAMP	The timestamp when the query was executed.
index_used	BOOLEAN	Specifies whether the query used the index.

Loading

sys.load_errors

The load_errors table shows information for load errors in the system.

Column Name	Column Type	Column Description
node_id	UUID	The unique identifier of the Loader Node.
table_id	UUID	If the event is specific to a table, the unique identifier of the table.
error_message	VARCHAR	The message that provides more details about the error.
error_timestamp	TIMESTAMP	The timestamp when this error occurred.
load_id	UUID	If the event is specific to a pipeline or query, the Universally Unique IDentifier (UUID) of the pipeline or query.

sys.load_events

The load_events table shows information for the load events in the system.

Column Name	Column Type	Column Description
node_id	UUID	The unique identifier of the Loader Node.
table_id	UUID	If the event is specific to a table, the unique identifier of the table.
event_type	VARCHAR	The type of the event.
event_timestamp	TIMESTAMP	The timestamp when this event occurred.
load_id	UUID	If the event is specific to a pipeline or query, the unique identifier of the pipeline or query.
batch_id	UUID	If the event is specific to segment generation, this column is the unique identifier for the batch of pages specified by the system for segment generation.
event_message	VARCHAR	The message that provides more details about the event.

sys.load_metrics

The load_metrics table shows information for the load metrics in the system.

Column Name	Column Type	Column Description
node_id	UUID	The Universally Unique IDentifier (UUID) of the Loader Node (sys.nodes).
table_id	UUID	If the metric is specific to a table, the UUID of the table (sys.tables).
load_id	UUID	If the metric is specific to a data pipeline or query, the UUID of the data pipeline or query.
name	VARCHAR	The type of the metric.
component	VARCHAR	The component of the metric.
value	BIGINT	The value of the metric.
updated_at	TIMESTAMP	Timestamp that represents when this metric was updated.

sys.segment_batch_events

This table contains information about segment batches during segment generation.

Column Name	Column Type	Column Description
batch_id	UUID	The Universally Unique IDentifier (UUID) for the batch.
table_id	UUID	The UUID for the table.
num_partitions	BIGINT	The number of partitions in this batch.
bucket_min	BIGINT	The size, in bytes, of the smallest bucket in this batch.
bucket_max	BIGINT	The size, in bytes, of the largest bucket in this batch.
scope_id	UUID	The UUID for the scope of this batch.
batch_start_time	TIMESTAMP	The start time of the batch analysis.
batch_end_time	TIMESTAMP	The end time of the batch analysis, including completion and partitioning.
batch_flush_reason	VARCHAR	The reason for the trigger of this batch analysis.
size	BIGINT	The total size, in bytes, of this batch.

sys.segment_generation_events

This table contains information about individual segments during segment generation.

Column Name	Column Type	Column Description
batch_id	UUID	The Universally Unique IDentifier (UUID) for the batch.
table_id	UUID	The UUID for the table.
ida_offset	BIGINT	The index number of the partition in this batch.
start_time	TIMESTAMP	The start time of segment analysis.
end_time	TIMESTAMP	The end time of segment analysis, including completion and partitioning.
size	BIGINT	The total size, in bytes, of this segment.
num_rows	BIGINT	The total number of rows in this segment.

Machine Learning

sys.association_rules_models

This table contains the name of the snapshot table for an association rules model.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
num_arguments	INT	Number of arguments in the machine learning model data.
table_name	CHAR	Name of the snapshot table.

sys.bagging_models

This table contains all model parameters for Bagging Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
aggregate_expression	CHAR	The vote expression that retrieves information from the child models.
num_features	INT	Number of arguments in the machine learning model data.
num_children	INT	Number of child models used by the bagging model.
accuracy	DOUBLE	Fraction of training data that the bagging model correctly classifies.
classes	ARRAY(CHAR)	Distinct output classes.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.
per_class_auc_roc	ARRAY(DOUBLE)	Areas under the Receiver Operating Characteristic (ROC) curve for each class, calculated using the One-vs-Rest (OVR) methodology.
macro_auc_roc	DOUBLE	Average area under the ROC curve across all classes.
weighted_auc_roc	DOUBLE	Average area under the ROC curve across all classes, weighted by class count.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.

sys.boosting_models

This table contains all model parameters for Boosting Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
expression	CHAR	The SQL expression that defines the boosting model.
num_arguments	INT	Number of arguments in the machine learning model data.
learning_rate	DOUBLE	Learning rate.
loss_function	CHAR	Loss function.
initial_predictions	ARRAY(DOUBLE)	Initial predictions.
accuracy	DOUBLE	Fraction of training data that the boosting model correctly classifies.
classes	ARRAY(CHAR)	Distinct output classes.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.
per_class_auc_roc	ARRAY(DOUBLE)	Areas under the Receiver Operating Characteristic (ROC) curve for each class, calculated using the One-vs-Rest (OVR) methodology.
macro_auc_roc	DOUBLE	Average area under the ROC curve across all classes.
weighted_auc_roc	DOUBLE	Average area under the ROC curve across all classes, weighted by class count.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.

sys.dbscan_models

This table contains all model parameters for Density-Based Spatial Clustering of Applications with Noise (DBSCAN) models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
num_features	INT	Number of feature columns the DBSCAN model trains on.
epsilon	DOUBLE	Distance threshold for the eps-neighborhood.
min_pts	INT	Minimum neighbors (including self) required for core-point status.
metric	CHAR	Distance metric with values: ‘euclidean’ (Euclidean), ‘cosine’ (Cosine), or ‘manhattan’ (Manhattan).
k_neighbors	INT	Internal KNN graph k that the model uses at training time.
insert_batch_size	LONG	Chunk size the database uses when materializing the cores table at the end of training: each multi-row `INSERT INTO temp.<clmap> VALUES (...)` statement holds at most this many `(row_id, cluster_id)` pairs. The default value is 131072 (result of 128 * 1024).
num_clusters	INT	Number of clusters the model discovers. Excludes the noise label -1.
num_noise_points	LONG	Count of noise rows (cluster identifier -1) the model observes during training.
num_core_points	LONG	Count of core points the model classifies.
num_border_points	LONG	Count of border points the model observes during training.
cluster_sizes	ARRAY(LONG)	Per-cluster training row counts (excludes noise). Index is the cluster identifier.
feature_names	ARRAY(CHAR)	Per-feature column names the model captures at training time. This value is empty when the source `SELECT` SQL statement does not name the column.
silhouette	DOUBLE	Silhouette coefficient over the assigned (non-noise) training rows. A higher value is better. This value is NULL when the model forms fewer than two clusters.
davies_bouldin	DOUBLE	Davies-Bouldin index. A lower value is better. This value is NULL when the model forms fewer than two clusters.
calinski_harabasz	DOUBLE	Calinski-Harabasz score. A higher value is better. This value is NULL when the model forms fewer than two clusters.
wcss	DOUBLE	Sum of squared distances from each assigned (non-noise) row to its cluster centroid in the normalized feature space.

sys.decision_tree_models

This table contains all model parameters for Decision Tree Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
case_statement	CHAR	The case statement that defines the decision tree.
case_statement_confidence	CHAR	The case statement that defines the decision tree and returns confidences.
num_arguments	INT	Number of arguments in the machine learning model data.
argument_types	ARRAY(CHAR)	Argument type names.
output_type	CHAR	Output type name.
disable_jit	BOOLEAN	Determines whether to disable JIT for the decision tree. Set this value to `true` to disable JIT. The default value is `false`.
accuracy	DOUBLE	Fraction of training data that the decision tree model correctly classifies.
classes	ARRAY(CHAR)	Distinct output classes.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.
per_class_auc_roc	ARRAY(DOUBLE)	Areas under the Receiver Operating Characteristic (ROC) curve for each class, calculated using the One-vs-Rest (OVR) methodology.
macro_auc_roc	DOUBLE	Average area under the ROC curve across all classes.
weighted_auc_roc	DOUBLE	Average area under the ROC curve across all classes, weighted by class count.
feature_importances	ARRAY(DOUBLE)	Feature importances for each input feature.

sys.feature_selection_models

This table contains parameters of every Backward Feature Elimination and Forward Feature Selection model in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
model_kind	CHAR	Search direction of the wrapper model with values: `BACKWARD` (backward) or `FORWARD` (forward).
inner_model_name	CHAR	Qualified name of the surviving inner model.
inner_model_type	CHAR	Model type of the surviving inner model.
evaluation_metric	CHAR	The criterion that drives the search loop.
metric_direction	CHAR	Direction of the metric with values: `higher_is_better` (higher is better) or `lower_is_better` (lower is better).
num_features_total	INT	Total number of features the wrapper model considers.
num_features_selected	INT	Number of features in the final selected subset.
selected_column_indices	ARRAY(INT)	Indexes of the selected parent input columns. The index starts at 1.
reference_score	DOUBLE	Score of the full-feature reference model. This value is NULL if no reference model was trained.
final_score	DOUBLE	Score of the final selected subset on the configured `evaluationMetric` option.
num_iterations_run	INT	Number of search rounds executed.
num_trials_run	INT	Total number of trial models trained across all rounds.
accuracy	DOUBLE	Classification accuracy of the final selected model. This value is NULL for regression and when you do not request metrics.
f1_macro	DOUBLE	Macro-averaged F1 score of the final selected model. This value is NULL for regression and when you do not request metrics.
auc_roc_macro	DOUBLE	Macro-averaged AUC ROC of the final selected model. This value is NULL for regression and when you do not request metrics.
mcc	DOUBLE	Matthews correlation coefficient of the final selected model. This value is NULL for regression and when you do not request metrics.
rmse	DOUBLE	Root mean squared error of the final selected model. This value is NULL for classification and when you do not request metrics.
mae	DOUBLE	Mean absolute error of the final selected model. This value is NULL for classification and when you do not request metrics.
r_squared	DOUBLE	R-squared of the final selected model. This value is NULL for classification and when you do not request metrics.

sys.feature_selection_steps

This table contains the per-accepted-move trace for every Backward Feature Elimination and Forward Feature Selection model in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the parent feature-selection machine learning model (sys.feature_selection_models).
step_index	INT	Zero-based index of this step in the accepted-move trace.
move_kind	CHAR	Specifies the action of the move with values: `ADD`, `REMOVE`, or `INITIAL`.
moved_column	INT	Index of the column added or removed by this step. The index starts at 1. This value is NULL for the initial step.
selected_after	ARRAY(INT)	Indexes of the columns selected after this step. Index starts at 1.
score_after	DOUBLE	Score of the selected subset after this step on the configured `evaluationMetric` option.
trials_in_round	INT	Number of trial models trained during the round that produced this step.

sys.feedforward_network_models

This table contains all model parameters for Feedforward Neural Network Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the feedforward network model.
num_arguments	INT	Number of arguments in the machine learning model data.
average_loss	DOUBLE	Average value of the loss function.

sys.gaussian_discriminant_analysis_models

This table contains all model parameters for Gaussian Discriminant Analysis (GDA) classification models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
num_features	INT	Number of features the GDA model trains on.
classes	ARRAY(CHAR)	Class labels in the order they appear in the per-class model arrays.
class_means	ARRAY(ARRAY(DOUBLE))	Per-class mean vectors in the normalized feature space of the model. The outer index is the class and the inner index is the feature.
log_dets	ARRAY(DOUBLE)	Per-class log determinant of the (regularized) covariance matrix in the normalized feature space.
log_priors	ARRAY(DOUBLE)	Per-class log prior probability the model uses during inference.
feature_means	ARRAY(DOUBLE)	Means the model uses at training time to normalize each input feature. When you set the `normalize` option to `false`, the values are identity values.
feature_stddevs	ARRAY(DOUBLE)	Standard deviations the model uses at training time to normalize each input feature. When you set the `normalize` option to `false`, the values are identity values.
accuracy	DOUBLE	Fraction of validation rows the model classifies correctly.
metrics_classes	ARRAY(CHAR)	Distinct output classes from metrics computation.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element at the ith row and jth column denotes the number of actual class i predicted to be class j.
per_class_precision	ARRAY(DOUBLE)	Precision per class using One-vs-Rest (OvR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recall per class using OvR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 per class using OvR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 over all classes.
weighted_precision	DOUBLE	Average precision over all classes weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes weighted by class count.
weighted_f1	DOUBLE	Average F1 over all classes weighted by class count.
mcc	DOUBLE	Multi-class Matthews correlation coefficient.
per_class_auc_roc	ARRAY(DOUBLE)	Per-class AUC ROC the model computes against shifted-softmax probabilities.
macro_auc_roc	DOUBLE	Unweighted average AUC ROC across classes.
weighted_auc_roc	DOUBLE	Class-count-weighted average AUC ROC across classes.

sys.gaussian_mixture_models

This table contains all model parameters for Gaussian Mixture Models.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the Gaussian Mixture Model.
num_distributions	INT	Number of distributions in the Gaussian Mixture Model.
average_loss	DOUBLE	Average value of the loss function.
log_likelihood	DOUBLE	Log-likelihood score of the Gaussian Mixture Model.

sys.gradient_boosted_trees_models

This table contains all model parameters for Gradient Boosted Trees Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
expression	CHAR	The SQL expression that defines the gradient boosted tree.
num_arguments	INT	Number of arguments in the machine learning model data.
initial_predictions	ARRAY(DOUBLE)	Initial predictions.
learning_rate	DOUBLE	Learning rate.
loss_function	CHAR	Loss function.
feature_importances	ARRAY(DOUBLE)	Feature importances for each input feature.
accuracy	DOUBLE	Fraction of training data that the gradient boosted trees model correctly classifies.
classes	ARRAY(CHAR)	Distinct output classes.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.
per_class_auc_roc	ARRAY(DOUBLE)	Areas under the Receiver Operating Characteristic (ROC) curve for each class, calculated using the One-vs-Rest (OVR) methodology.
macro_auc_roc	DOUBLE	Average area under the ROC curve across all classes.
weighted_auc_roc	DOUBLE	Average area under the ROC curve across all classes, weighted by class count.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.

sys.k_means_models

This table contains all model parameters for K-Means Clustering Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
centroids	ARRAY(DOUBLE)	Values that represent the centroid values of each cluster.
num_arguments	INT	Number of arguments in the machine learning model data.
means	ARRAY(DOUBLE)	Means of each of the feature columns in the data.
standard_deviations	ARRAY(DOUBLE)	Standard deviations of each of the feature columns in the data.
wcss	DOUBLE	Within-cluster sum of squares score for the model.

sys.k_nearest_neighbors_models

This table contains the number of arguments and the target table for K-Nearest Neighbors Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
num_arguments	INT	Number of arguments in the machine learning model data.
table_name	CHAR	Name of the target table for the K Nearest Neighbors (KNN) model.
means	ARRAY(DOUBLE)	Means of each of the feature columns in the data.
standard_deviations	ARRAY(DOUBLE)	Standard deviations of each of the feature columns in the data.
accuracy	DOUBLE	Fraction of training data that the KNN model correctly classifies.
classes	ARRAY(CHAR)	Distinct output classes.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.

sys.linear_combination_regression_models

This table contains all model parameters for Linear Combination Regression Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of model coefficients associated with the regression.
y_intercept	DOUBLE	y-intercept of the regression.
num_arguments	INT	Number of arguments in the machine learning model data.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.
aic	DOUBLE	Akaike information criterion value.
bic	DOUBLE	Bayesian information criterion value.
optimizer_used	CHAR	Optimizer used during training: ‘closed_form’, ‘fista’, or ‘irls_huber’. NULL for models trained before this column was added.
iterations_used	INT	Number of iterations the iterative optimizer used during training. NULL for the closed-form path or for models trained before this column was added.
converged	BOOLEAN	Whether the iterative optimizer converged before reaching its iteration cap. NULL for the closed-form path or for models trained before this column was added.
huber_outlier_count	LONG	Number of training rows whose final residual exceeded the effective Huber threshold (delta * sigma). Populated only when lossFunction=‘huber’.

sys.linear_discriminant_analysis_models

This table contains all model parameters for Linear Discriminant Analysis (LDA) Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the LDA model.
num_features	INT	Number of features returned by the LDA model.
importance	ARRAY(DOUBLE)	Importance of each of the output features that represent the data.

sys.logistic_regression_models

This table contains all model parameters for Logistic Regression Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the logistic regression model.
num_arguments	INT	Number of arguments in the machine learning model data.
zero_case	CHAR	The case where the logistic regression uses a value of 0.
one_case	CHAR	The case where the logistic regression uses a value of 1.
classes	ARRAY(CHAR)	Target classes for the logistic regression.
accuracy	DOUBLE	Fraction of training data that the logistic regression model correctly classifies.
metrics_classes	ARRAY(CHAR)	Distinct output classes from metrics computation.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.
per_class_auc_roc	ARRAY(DOUBLE)	Areas under the Receiver Operating Characteristic (ROC) curve for each class, calculated using the One-vs-Rest (OVR) methodology.
macro_auc_roc	DOUBLE	Average area under the ROC curve across all classes.
weighted_auc_roc	DOUBLE	Average area under the ROC curve across all classes, weighted by class count.

sys.machine_learning_model_options

This table contains options defined for each machine learning model. The database stores these options as key-value pairs.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	UUID of the machine learning model (sys.machine_learning_models).
machine_learning_model_key	CHAR	Machine learning model configuration option key.
machine_learning_model_value	CHAR	Machine learning model configuration option value.

sys.machine_learning_models

This table contains the machine learning models that are defined in the database.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the machine learning model.
name	CHAR	Name of the machine learning model.
schema	CHAR	Schema of the machine learning model.
machine_learning_model_type	CHAR	Type of the machine learning model.
database_id	UUID	UUID of the database that contains the machine learning model definition (sys.databases).
on_select	CHAR	The SELECT SQL statement used to create this model.
creator_id	UUID	The UUID of the user or group who created the machine learning model (sys.users/sys.groups).

sys.mlp_models

This table contains all model parameters for Multi-Layer Perceptron (MLP) Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the MLP model.
num_arguments	INT	Number of arguments in the machine learning model data.
average_loss	DOUBLE	Average value of the loss function.

sys.multiple_linear_regression_models

This table contains the y-intercept and coefficient of determination values for the multiple linear regression machine learning models.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
y_intercept	DOUBLE	y-intercept of the regression.
slopes	ARRAY(DOUBLE)	Slopes for each variable in the regression.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.
aic	DOUBLE	Akaike information criterion value.
bic	DOUBLE	Bayesian information criterion value.
optimizer_used	CHAR	Optimizer used during training: ‘closed_form’, ‘fista’, or ‘irls_huber’. NULL for models trained before this column was added.
iterations_used	INT	Number of iterations the iterative optimizer used during training. NULL for the closed-form path or for models trained before this column was added.
converged	BOOLEAN	Whether the iterative optimizer converged before reaching its iteration cap. NULL for the closed-form path or for models trained before this column was added.
huber_outlier_count	LONG	Number of training rows whose final residual exceeded the effective Huber threshold (delta * sigma). Populated only when lossFunction=‘huber’.

sys.naive_bayes_models

This table contains all model parameters for Naive Bayes Classification Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
num_arguments	INT	Number of arguments in the machine learning model data.
result_probability_table	CHAR	Name of the internal result probability table.
feature_result_matrix_table	CHAR	Name of the internal feature result matrix table.
accuracy	DOUBLE	Fraction of training data that the Naive Bayes model correctly classifies.
classes	ARRAY(CHAR)	Distinct output classes from metrics computation.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.

sys.nonlinear_regression_models

This table contains all model parameters for Nonlinear Regression Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the nonlinear regression model.
num_arguments	INT	Number of arguments in the machine learning model data.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.
aic	DOUBLE	Akaike information criterion value.
bic	DOUBLE	Bayesian information criterion value.
constraints	CHAR	The SQL Boolean predicate over `a1..aN` that the model uses for training. This value is NULL when you create the model without this option set to the `OPTIONS` value, or for models the system trains before the addition of this column.
accepted_steps	LONG	Count of optimizer steps that satisfy the constraint predicate during training. This value is NULL when you create the model without this option set to the `OPTIONS` value.
rejected_steps	LONG	Count of optimizer steps that violate the constraint predicate and that the model rejects during training. This value is NULL when you create the model without this option set to the `OPTIONS` value.
final_lambda	DOUBLE	Final Levenberg-Marquardt damping factor at the training termination. This value is NULL for Adam-trained models (set the `lossFunction` option to a function that is different from the default least-squares form) or when you create the model without this option set to the `OPTIONS` value.
constraint_infeasible	BOOLEAN	Determines whether the constraint is infeasible. This value is `true` if training terminates because the optimizer hit the number of consecutive constraint-rejected steps that you specify using the `maxConsecutiveRejections` option without an accept (constraint surface is unreachable from the chosen starting point). This value is NULL when you create the model without this option set to the `OPTIONS` value.

sys.polynomial_regression_models

This table contains all model parameters for Polynomial Regression Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of model coefficients associated with the polynomial regression model.
y_intercept	DOUBLE	y-intercept of the regression.
num_arguments	INT	Number of arguments in the machine learning model data.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.
aic	DOUBLE	Akaike information criterion value.
bic	DOUBLE	Bayesian information criterion value.
optimizer_used	CHAR	Optimizer used during training: ‘closed_form’, ‘fista’, or ‘irls_huber’. NULL for models trained before this column was added.
iterations_used	INT	Number of iterations the iterative optimizer used during training. NULL for the closed-form path or for models trained before this column was added.
converged	BOOLEAN	Whether the iterative optimizer converged before reaching its iteration cap. NULL for the closed-form path or for models trained before this column was added.
huber_outlier_count	LONG	Number of training rows whose final residual exceeded the effective Huber threshold (delta * sigma). Populated only when lossFunction=‘huber’.

sys.principal_component_analysis_models

This table contains all model parameters for Principal Component Analysis (PCA) Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the PCA model.
num_features	INT	Number of features returned by the PCA model.
importance	ARRAY(DOUBLE)	Importance of each of the output features that represents the data.
means	ARRAY(DOUBLE)	Means of each of the feature columns in the data.
standard_deviations	ARRAY(DOUBLE)	Standard deviations of each of the feature columns in the data.
num_components	INT	Number of principal components retained by the model.
whiten	BOOLEAN	Whether the model scales component scores by the inverse square root of eigenvalues.
eigenvalues	ARRAY(DOUBLE)	Eigenvalues for each principal component.

sys.random_forest_models

This table contains all model parameters for Random Forest Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
vote_expression	CHAR	The vote expression that retrieves information from the child decision trees.
vote_expression_confidence	CHAR	The vote expression that retrieves information from the child decision trees and returns confidences.
num_arguments	INT	Number of arguments in the machine learning model data.
num_children	INT	Number of child trees in the forest.
feature_importances	ARRAY(DOUBLE)	Feature importances for each input feature.
oob_score	DOUBLE	Out-of-bag score of the random forest model.
accuracy	DOUBLE	Fraction of training data that the random forest model correctly classifies.
classes	ARRAY(CHAR)	Distinct output classes from metrics computation.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.
per_class_auc_roc	ARRAY(DOUBLE)	Areas under the Receiver Operating Characteristic (ROC) curve for each class, calculated using the One-vs-Rest (OVR) methodology.
macro_auc_roc	DOUBLE	Average area under the ROC curve across all classes.
weighted_auc_roc	DOUBLE	Average area under the ROC curve across all classes, weighted by class count.

sys.regression_tree_models

This table contains all model parameters for Regression Tree Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
case_statement	CHAR	The case statement that defines the regression tree.
num_arguments	INT	Number of arguments in the machine learning model data.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.
feature_importances	ARRAY(DOUBLE)	Feature importances for each input feature.

sys.simple_linear_regression_models

This table contains the slope, y-intercept, and coefficient of determination values for single linear regression machine learning models.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
slope	DOUBLE	Slope of the regression.
y_intercept	DOUBLE	The y-intercept of the regression.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.
aic	DOUBLE	Akaike information criterion value.
bic	DOUBLE	Bayesian information criterion value.

sys.stacking_models

This table contains all model parameters for Stacking Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
num_arguments	INT	Number of arguments in the machine learning model data.
accuracy	DOUBLE	Fraction of training data that the stacking model correctly classifies.
classes	ARRAY(CHAR)	Distinct output classes from metrics computation.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.
r2	DOUBLE	R^2 value.
adjusted_r2	DOUBLE	Adjusted R^2 value.
rmse	DOUBLE	Root mean square error.
mae	DOUBLE	Mean absolute error.
mape	DOUBLE	Mean absolute percentage error.

sys.support_vector_machine_models

This table contains all model parameters for Support Vector Machine (SVM) Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the SVM model.
num_arguments	INT	Number of arguments in the machine learning model data.
negative_case	CHAR	Case when the SVM model classifies data as negative.
positive_case	CHAR	Case when the SVM model classifies data as positive.
classes	ARRAY(CHAR)	Target classes for the SVM model.
accuracy	DOUBLE	Fraction of training data that the SVM model correctly classifies.
metrics_classes	ARRAY(CHAR)	Distinct output classes from metrics computation.
confusion_matrix	ARRAY(ARRAY(LONG))	Confusion matrix, where the element ith row and jth column denotes the number of actual class i that were predicted to be j.
per_class_precision	ARRAY(DOUBLE)	Precisions per class using One-vs-Rest (OVR) calculations.
per_class_recall	ARRAY(DOUBLE)	Recalls per class using OVR calculations.
per_class_f1	ARRAY(DOUBLE)	F1 scores per class using OVR calculations.
macro_precision	DOUBLE	Average precision over all classes.
macro_recall	DOUBLE	Average recall over all classes.
macro_f1	DOUBLE	Average F1 score over all classes.
weighted_precision	DOUBLE	Average precision over all classes, weighted by class count.
weighted_recall	DOUBLE	Average recall over all classes, weighted by class count.
weighted_f1	DOUBLE	Average F1 score over all classes, weighted by class count.
mcc	DOUBLE	Matthews correlation coefficient.
per_class_auc_roc	ARRAY(DOUBLE)	Areas under the Receiver Operating Characteristic (ROC) curve for each class, calculated using the One-vs-Rest (OVR) methodology.
macro_auc_roc	DOUBLE	Average area under the ROC curve across all classes.
weighted_auc_roc	DOUBLE	Average area under the ROC curve across all classes, weighted by class count.

sys.survival_regression_models

This table contains all model parameters for Survival Regression Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
family	CHAR	Survival regression family with supported values: ‘cox’ (Cox proportional hazards (PH)) or ‘aft’ (accelerated failure time).
aft_distribution	CHAR	Accelerated Failure Time (AFT) distribution with supported values: ‘weibull’, ‘lognormal’, or ‘loglogistic’. When you set the `family` option to ‘cox’, this value is empty.
ties	CHAR	Cox proportional hazards ties handling with supported values: ‘efron’ or ‘breslow’. When you set the `family` option to ‘aft’, this value is empty.
coefficients	ARRAY(DOUBLE)	Linear-predictor coefficients (one per feature). This value uses the partial-likelihood coefficients for Cox PH, and the log-time location coefficients for AFT.
aft_intercept	DOUBLE	AFT intercept (mu). When you set the `family` option to ‘cox’, this value is NULL.
aft_scale	DOUBLE	AFT scale parameter (sigma). When you set the `family` option to ‘cox’, this value is NULL.
final_log_likelihood	DOUBLE	Partial log-likelihood at convergence when you set the `family` option to ‘cox’, or full log-likelihood when you set that option to ‘aft’.
num_arguments	INT	Number of feature columns in the machine learning model data.
concordance_index	DOUBLE	The Concordance Index. If the model does not compute metrics, this value is NULL.

sys.undersampling_models

This table contains the parameters and the output table name for an Undersampling Preprocessing model. The output table named in the output_table column contains sampled rows that downstream classifiers can train against.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
class_column	CHAR	Name of the column with the value that defines the class for sampling.
strategy	CHAR	Sampling strategy with values: ‘random’ or ‘stratified’.
sampling_ratio	DOUBLE	Final majority or minority row count ratio relative to the target class.
output_table	CHAR	Name of the output table in the `temp` schema. Read this table to access the sampled rows. The system drops this table when you execute the `DROP MLMODEL` SQL statement.
target_class	CHAR	Class label that defines the target row count, specified as either the smallest class (when targetClass=‘auto’) or the explicitly named class.
stratum_column	CHAR	Name of the column defining strata when you set the `strategy` option to ‘stratified’. When you set the `strategy` option to ‘random’, this value is empty.

sys.vector_autoregression_models

This table contains all model parameters for Vector Autoregression Models in the system.

Column Name	Column Type	Column Description
machine_learning_model_id	UUID	Universally Unique IDentifier (UUID) of the machine learning model (sys.machine_learning_models).
coefficients	ARRAY(DOUBLE)	List of coefficients associated with the vector autoregression model.
coefficient_of_determination	DOUBLE	R^2 value of the autoregression, if calculated, otherwise NULL.
optimizer_used	CHAR	Optimizer used during training: ‘closed_form’ or ‘fista’. NULL for models trained before this column was added.
iterations_used	INT	Number of FISTA iterations the optimizer used during training. NULL for the closed-form path or for models trained before this column was added.
converged	BOOLEAN	Whether the FISTA optimizer converged before reaching its iteration cap. NULL for the closed-form path or for models trained before this column was added.

Model Context Protocol

sys.mcp_tools

This table contains all of the Model Context Protocol (MCP) tools in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the MCP tool.
database_id	UUID	Universally Unique IDentifier (UUID) of the database containing the MCP tool. This value is NULL for system tools.
schema	CHAR	Name of the schema.
name	CHAR	Name of the MCP tool.
argument_names	ARRAY(CHAR)	The argument names of the MCP tool, specified as an array.
argument_types	ARRAY(CHAR)	The argument types of the MCP tool, specified as an array.
argument_nullability	ARRAY(BOOLEAN)	Whether the argument types for the MCP tool are nullable, specified as an array.
language	CHAR	Programming language of the MCP tool.
description	CHAR	Description of the MCP tool.
toolset	CHAR	Toolset where the MCP tool belongs.
definition	CHAR	Function definition of the MCP tool.
tool_type	CHAR	Type of the MCP tool with values: `MCP TOOL` for user-defined tools and `SYSTEM MCP TOOL` for system tools.
creator_id	UUID	Universally Unique IDentifier (UUID) of the user who created the MCP tool. This value is NULL for system tools.
created_at	TIMESTAMP	Timestamp that represents the MCP tool was created. This value is NULL for system tools.

Monitoring

sys.active_operator_instances

This table contains information about operator instances that are active in the database.

Column Name	Column Type	Column Description
database_name	CHAR	The name of the database associated with the query to which the operator belongs.
user_name	CHAR	The username associated with the query to which the operator belongs.
query_id	UUID	A unique identifier of the query to which the operator belongs.
operator_id	UUID	A unique identifier of the finalized operator.
node_id	UUID	The unique identifier of the node that stores the information about this operator.
silo_id	LONG	The unique identifier of the silo that stores the information about this operator.
core_id	LONG	The unique identifier of the virtual machine core that stores the information about this operator.
op_runtime_id	LONG	The unique identifier for the operator instance that the system assigns at runtime.
op_instance_type	CHAR	The name of the type for this operator instance.
bloom_filtered_rows	LONG	The rows filtered by Bloom filters.
rows_received	LONG	The number of rows fetched by the operator instance.
rows_processed	LONG	The number of rows processed by this operator instance during execution.
rows_emitted	LONG	The number of rows returned by the operator instance.
blocks_emitted	LONG	The number of data blocks returned by the operator instance.
first_block_receive_time	TIMESTAMP	The timestamp that specifies the return of the first data block.
first_block_emitted_time	TIMESTAMP	The timestamp that specifies the return of the first data block.
num_cycles	LONG	The number of cycles that run on this operator instance.
num_oom_cycles	LONG	The number of run cycles that an operator received for processing out-of-memory issues.
num_failed_oom_cycles	LONG	The number of out-of-memory cycles received that failed to do any work.
num_no_work_oom_cycles	LONG	The number of out-of-memory cycles received that reported they had no work to do.
run_count	LONG	The number of times that the scheduler has reviewed this operator and `isRunnable` is set to `true`.
no_work_cycles	LONG	The number of times that the read and write cycle executed and performed no work.
backed_up_cycles	LONG	The number of times a cycle stopped because a backup of the parent exists.
leaf_branch_blocked_cycles	LONG	The number of times a cycle stopped on a leaf operator because of multi-child scheduling heuristics.
num_eof_cycles	LONG	The number of end-of-file cycles received.
normal_run_time_in_ms	DOUBLE	The time, in milliseconds, of the normal run for the operator instance.
dispatch_queue_time_in_ms	DOUBLE	The time spent, in milliseconds, processing incoming message events for the operator instance.
oom_run_time_in_ms	DOUBLE	The time, in milliseconds, of the out-of-memory issue for the operator instance.
max_hp_mem_usage	LONG	The maximum heap memory usage by the operator instance.
num_blocks_written	LONG	The number of data blocks written to temporary disk.
num_blocks_read	LONG	The number of data blocks read from temporary disk.
num_fragments_written	LONG	The number of written memory fragments.
num_fragments_read	LONG	The number of read memory fragments.
cycles_over_1_sec	LONG	The count of cycles that ran over 1 second for this operator instance.
cycles_over_3_sec	LONG	The count of cycles that ran over 3 seconds for this operator instance.
long_dispatch_queue_events	LONG	The count of dispatch queue events processed by the operator with a runtime greater than 0.5 seconds.
total_bin_stream_allocated_bytes	LONG	The total amount of memory allocated in output blocks for variable-length column data.
total_bin_stream_used_bytes	LONG	The total amount of memory used in output blocks for variable-length column data.
total_col_stream_allocated_bytes	LONG	The total amount of memory allocated in output blocks for fixed-length column data.
total_col_stream_used_bytes	LONG	The total amount of memory used in output blocks for fixed-length column data.
blocked_received	LONG	The total number of input data blocks received by this operator instance.
blocked_created	LONG	The total number of non-trivial output data blocks created by this operator instance.
received_block_bin_stream_unused_space_bytes	LONG	The total amount of wasted space across all binary streams in received data blocks (includes over-allocations as well as projected columns and dead memory). This number is NULL unless configuration parameters allow recording this statistic for each block because the calculation is expensive.
received_block_col_stream_unused_space_bytes	LONG	The total amount of wasted space across all column streams in received data blocks (includes over-allocations as well as projected columns and dead memory). This number is NULL unless configuration parameters allow recording this statistic for each block because the calculation is expensive.
mandatory_alloc_attempts	LONG	The count of cycles where the scheduler enabled mandatory allocation.
mandatory_alloc_success_count	LONG	The count of cycles where the scheduler enabled mandatory allocation and this operator broke an out-of-memory deadlock.
max_num_pending_blocks	LONG	The maximum number of pending blocks across all partitions during the lifetime of this operator.
max_num_consecutive_backup_cycles	LONG	The maximum number of consecutive scheduler cycles where this operator has a backup. The scheduler can use this information to help detect certain resource deadlocks.
longest_time_since_unblocked_cycle	LONG	The maximum time between successful attempts to run this operator from the scheduler. The scheduler can use this information to help detect certain resource deadlocks.
compiled_level	LONG	The level within a hierarchy of distinct operators to execute. A single plan operator can be compiled into this hierarchy.
additional_json	CHAR	Additional statistics for an operator instance in JSON format.
num_bloom_filters_sent	LONG	The number of Bloom filters the operator sent to its peers.
num_bloom_filters_received	LONG	The number of Bloom filters the operator received from its peers.
num_key_lists_sent	LONG	The number of potential index join key lists the operator sent to its peers.
num_key_lists_received	LONG	The number of potential index join key lists the operator received from its peers.
max_hash_table_rows	LONG	The maximum number of rows present in any hash table used by this operator instance.
plan_operator_estimated_output_cardinality	LONG	The number of rows the optimizer estimated for this plan operator to emit.
num_delayed_exceptions_generated	LONG	The number of rows this operator generated with delayed exception information.
num_pending_blocks_spilled	LONG	The number of queued pending data blocks this operator spilled to disk.
max_normal_cycle_time_ms	DOUBLE	The time, in milliseconds, of the longest normal run cycle for the operator instance.
max_oom_cycle_time_ms	DOUBLE	The time, in milliseconds, of the longest out-of-memory run cycle for the operator instance.
num_rows_received_per_plan_child	ARRAY(LONG)	The number of rows this operator received from each of its children in the plan.
initialization_time	TIMESTAMP	The timestamp that represents when the operator began work.
num_cycles_blocked_by_temp_disk	LONG	The count of cycles that the operator spent waiting for input and output from the temporary disk.
num_cycles_blocked_by_external_pending_blocks	LONG	The count of cycles that the operator spent waiting for input and output from external pending blocks.
blobstore_callback_time_ms	DOUBLE	The time, in milliseconds, the system spends processing callbacks for temporary disk input and output for the operator instance.
num_bounding_box_lists_sent	LONG	The number of geospatial bounding box list filters the operator sent to its peers.
num_bounding_box_lists_received	LONG	The number of geospatial bounding box list filters the operator received from its peers.
target_rows_per_cycle	LONG	The number of rows this operator expects to process in a single cycle without exceeding time limits.
applied_dynamic_filter_stats	CHAR	Additional statistics about dynamic pushdown filters in JSON format.
is_runnable	BOOLEAN	Whether the operator has available data to process.
current_num_pending_blocks	LONG	The current number of pending blocks across all partitions of this operator.

sys.active_query_scheduler_info

This table contains information about how the scheduler schedules each query to execute within the virtual machine (VM).

Column Name	Column Type	Column Description
query_id	UUID	The Universally Unique IDentifier (UUID) of the query that the scheduler currently schedules to execute.
node_id	UUID	The UUID of the node that executes this query.
silo_id	LONG	The identifier of the silo that executes this query on its associated node.
core_id	LONG	The identifier of the VM core that executes this query on its associated silo.
virtual_runtime	DOUBLE	A unitless fairness metric the scheduler uses to determine the query to schedule next. Comparisons of this value are only meaningful across concurrently executing queries with identical `node_id`, `silo_id`, `core_id`, and `current_priority` values.
current_priority	DOUBLE	The current scheduling priority of this query on its VM core.
total_cycles	LONG	The number of times this VM core has attempted to schedule this query.
successful_cycles	LONG	The number of times this VM core has attempted to schedule this query and completed some amount of work.
no_work_cycles	LONG	The number of times this VM core has attempted to schedule this query and no operators were prepared to do work.
failed_cycles	LONG	The number of times this VM core has attempted to schedule this query, some operators attempted to do work, and all operators failed to complete any amount of work.
failed_oom_cycles	LONG	The number of times this VM core has attempted to schedule this query, some operators attempted to do work, all operators failed to complete any amount of work, and at least one operator on the query has reported that it is out of memory.
num_greedy_cycles	LONG	The number of times this VM core has attempted to schedule this query, which is the exclusive highest priority query currently managed by this VM core.
num_try_chilled_ops_cycles	LONG	The number of times this VM core attempted to schedule this query and polled operators that have been marked as `chilled`, meaning they are unlikely to be prepared to do work.
num_successful_try_chilled_ops_cycles	LONG	The number of times this VM core attempted to schedule this query, polled operators that have been marked ‘chilled’, and successfully completed some amount of work.
total_op_inst_cycle_time	DOUBLE	The approximate total CPU time, in milliseconds, that the operators in this query have been scheduled to execute on this VM core.
total_op_inst_dispatch_queue_time	DOUBLE	The approximate total CPU time, in milliseconds, that operators associated with this query have spent processing incoming message events.
total_runtime	DOUBLE	The approximate execution time, in milliseconds, that this VM core has managed this query.
total_bytes_spilled	LONG	The approximate total temporary disk amount that the VM core uses during query execution.

sys.all_operator_instances

This table contains information about operator instances that have been created by the virtual machine.

Column Name	Column Type	Column Description
database_name	VARCHAR	The name of the database associated with the query to which the operator belongs.
user_name	VARCHAR	The username associated with the query to which the operator belongs.
query_id	UUID	A Universally Unique IDentifier (UUID) of the query to which the operator belongs.
operator_id	UUID	The UUID of the finalized operator.
node_id	UUID	The UUID of the node that stores information about this operator.
silo_id	BIGINT	The identifier of the silo that stores information about this operator.
core_id	BIGINT	The identifier of the virtual machine core that stores information about this operator.
op_runtime_id	BIGINT	The identifier for the operator instance that the system assigns at runtime.
op_instance_type	VARCHAR	The name of the type for this operator instance.
bloom_filtered_rows	BIGINT	The rows filtered by Bloom filters.
rows_received	BIGINT	The number of rows fetched by the operator instance.
rows_processed	BIGINT	The number of rows processed by this operator instance during execution.
rows_emitted	BIGINT	The number of rows returned by the operator instance.
blocks_emitted	BIGINT	The number of data blocks returned by the operator instance.
finalization_time	TIMESTAMP	The timestamp that represents when the operator completed all work.
first_block_receive_time	TIMESTAMP	The timestamp that specifies the receipt of the first data block.
first_block_emitted_time	TIMESTAMP	The timestamp that specifies the return of the first data block.
num_cycles	BIGINT	The number of cycles that run on this operator instance.
num_oom_cycles	BIGINT	The number of run cycles that an operator received for processing out-of-memory issues.
num_failed_oom_cycles	BIGINT	The number of out-of-memory cycles received that failed to do any work.
num_no_work_oom_cycles	BIGINT	The number of out-of-memory cycles received with no work to do.
run_count	BIGINT	The number of times that the scheduler has reviewed this operator and `isRunnable` is set to `true`.
no_work_cycles	BIGINT	The number of times that the read and write cycle executed and performed no work.
backed_up_cycles	BIGINT	The number of times a cycle stopped because a backup of the parent exists.
leaf_branch_blocked_cycles	BIGINT	The number of times a cycle stopped on a leaf operator because of multi-child scheduling heuristics.
num_eof_cycles	BIGINT	The number of end-of-file cycles received.
normal_run_time_in_ms	DOUBLE	The time, in milliseconds, of the normal run for the operator instance.
dispatch_queue_time_in_ms	DOUBLE	The time, in milliseconds, spent processing incoming message events for the operator instance.
oom_run_time_in_ms	DOUBLE	The time, in milliseconds, of the out-of-memory issue for the operator instance.
max_hp_mem_usage	BIGINT	The maximum heap memory usage by the operator instance.
num_blocks_written	BIGINT	The number of data blocks written to temporary disk.
num_blocks_read	BIGINT	The number of data blocks read from temporary disk.
num_fragments_written	BIGINT	The number of written memory fragments.
num_fragments_read	BIGINT	The number of read memory fragments.
cycles_over_1_sec	BIGINT	The count of cycles that ran over 1 second for this operator instance.
cycles_over_3_sec	BIGINT	The count of cycles that ran over 3 seconds for this operator instance.
long_dispatch_queue_events	BIGINT	The count of dispatch queue events processed by the operator with a runtime greater than 0.5 seconds.
total_bin_stream_allocated_bytes	BIGINT	The total amount of memory allocated in output blocks for variable-length column data.
total_bin_stream_used_bytes	BIGINT	The total amount of memory used in output blocks for variable-length column data.
total_col_stream_allocated_bytes	BIGINT	The total amount of memory allocated in output blocks for fixed-length column data.
total_col_stream_used_bytes	BIGINT	The total amount of memory used in output blocks for fixed-length column data.
blocked_received	BIGINT	The total number of input data blocks received by this operator instance.
blocked_created	BIGINT	The total number of non-trivial output data blocks created by this operator instance.
received_block_bin_stream_unused_space_bytes	BIGINT	The total amount of wasted space across all binary streams in received data blocks (includes over-allocations as well as projected columns and dead memory). This number is NULL unless configuration parameters allow recording this statistic for each block because the calculation is expensive.
received_block_col_stream_unused_space_bytes	BIGINT	The total amount of wasted space across all column streams in received data blocks (includes over-allocations as well as projected columns and dead memory). This number is NULL unless configuration parameters allow recording this statistic for each block because the calculation is expensive.
mandatory_alloc_attempts	BIGINT	The count of cycles where the scheduler enabled mandatory allocation.
mandatory_alloc_success_count	BIGINT	The count of cycles where the scheduler enabled mandatory allocation and this operator broke an out-of-memory deadlock.
max_num_pending_blocks	BIGINT	The maximum number of pending blocks across all partitions during the lifetime of this operator.
max_num_consecutive_backup_cycles	BIGINT	The maximum number of consecutive scheduler cycles where this operator has a backup. The scheduler can use this information to help detect certain resource deadlocks.
longest_time_since_unblocked_cycle	BIGINT	The maximum time between successful attempts to run this operator from the scheduler. The scheduler can use this information to help detect certain resource deadlocks.
compiled_level	BIGINT	The level within a hierarchy of distinct operators to execute. The system can compile a single plan operator into this hierarchy.
additional_json	VARCHAR	Additional statistics for an operator instance in JSON format.
num_bloom_filters_sent	BIGINT	The number of Bloom filters the operator sent to its peers.
num_bloom_filters_received	BIGINT	The number of Bloom filters the operator received from its peers.
num_key_lists_sent	BIGINT	The number of potential index join key lists the operator sent to its peers.
num_key_lists_received	BIGINT	The number of potential index join key lists the operator received from its peers.
max_hash_table_rows	BIGINT	The maximum number of rows present in any hash table used by this operator instance.
plan_operator_estimated_output_cardinality	BIGINT	The number of rows the optimizer estimated for this plan operator to emit.
num_delayed_exceptions_generated	BIGINT	The number of rows this operator generated with delayed exception information.
num_pending_blocks_spilled	BIGINT	The number of queued pending data blocks this operator spilled to disk.
max_normal_cycle_time_ms	DOUBLE	The time, in milliseconds, of the longest normal run cycle for the operator instance.
max_oom_cycle_time_ms	DOUBLE	The time, in milliseconds, of the longest out-of-memory run cycle for the operator instance.
num_rows_received_per_plan_child	BIGINT[]	The number of rows this operator received from each of its children in the plan.
initialization_time	TIMESTAMP	The timestamp that represents when the operator began work.
num_cycles_blocked_by_temp_disk	BIGINT	The count of cycles that the operator spent waiting for input and output from the temporary disk.
num_cycles_blocked_by_external_pending_blocks	BIGINT	The count of cycles that the operator spent waiting for input and output from external pending blocks.
blobstore_callback_time_ms	DOUBLE	The time, in milliseconds, the system spends processing callbacks for temporary disk input and output for the operator instance.
num_bounding_box_lists_sent	BIGINT	The number of geospatial bounding box list filters the operator sent to its peers.
num_bounding_box_lists_received	BIGINT	The number of geospatial bounding box list filters the operator received from its peers.
target_rows_per_cycle	BIGINT	The number of rows this operator expects to process in a single cycle without exceeding time limits.
applied_dynamic_filter_stats	VARCHAR	Additional statistics about dynamic pushdown filters in JSON format.
is_runnable	BOOLEAN	Whether the operator has available data to process.
current_num_pending_blocks	BIGINT	The current number of pending blocks across all partitions of this operator.
is_active	BOOLEAN	Whether the operator is currently active or has finalized.

sys.completed_operator_instances

This table contains information about operator instances that have finalized in the virtual machine.

Column Name	Column Type	Column Description
database_name	VARCHAR	The database associated with the query to which the operator belongs.
user_name	VARCHAR	The user associated with the query to which the operator belongs.
query_id	UUID	A unique identifier of the query to which the operator belongs.
operator_id	UUID	A unique identifier of the finalized operator.
node_id	UUID	The unique identifier of the node that stores the information about this operator.
silo_id	BIGINT	The unique identifier of the silo that stores the information about this operator.
core_id	BIGINT	The unique identifier of the virtual machine core that stores the information about this operator.
op_runtime_id	BIGINT	The unique identifier for the operator instance that the system assigns at runtime.
op_instance_type	VARCHAR	The name of the type for this operator instance.
bloom_filtered_rows	BIGINT	The rows filtered by Bloom filters.
rows_received	BIGINT	The number of rows fetched by the operator instance.
rows_processed	BIGINT	The number of rows processed by this operator instance during execution.
rows_emitted	BIGINT	The number of rows returned by the operator instance.
blocks_emitted	BIGINT	The number of data blocks returned by the operator instance.
finalization_time	TIMESTAMP	The timestamp that represents when the operator completed all work.
first_block_receive_time	TIMESTAMP	The timestamp that specifies the receipt of the first data block.
first_block_emitted_time	TIMESTAMP	The timestamp that specifies the return of the first data block.
num_cycles	BIGINT	The number of cycles that run on this operator instance.
num_oom_cycles	BIGINT	The number of run cycles that an operator received for processing out-of-memory issues.
num_failed_oom_cycles	BIGINT	The number of out-of-memory cycles received that failed to do any work.
num_no_work_oom_cycles	BIGINT	The number of out-of-memory cycles received that reported they had no work to do.
run_count	BIGINT	The number of times that the scheduler has reviewed this operator and `isRunnable` is set to `true`.
no_work_cycles	BIGINT	The number of times that the read and write cycle executed and performed no work.
backed_up_cycles	BIGINT	The number of times a cycle stopped because a backup of the parent exists.
leaf_branch_blocked_cycles	BIGINT	The number of times a cycle stopped on a leaf operator because of multi-child scheduling heuristics.
num_eof_cycles	BIGINT	The number of end-of-file cycles received.
normal_run_time_in_ms	DOUBLE	The time, in milliseconds, of the normal run for the operator instance.
dispatch_queue_time_in_ms	DOUBLE	The time spent, in milliseconds, processing incoming message events for the operator instance.
oom_run_time_in_ms	DOUBLE	The time, in milliseconds, of the out-of-memory issue for the operator instance.
max_hp_mem_usage	BIGINT	The maximum heap memory usage by the operator instance.
num_blocks_written	BIGINT	The number of data blocks written to temporary disk.
num_blocks_read	BIGINT	The number of data blocks read from temporary disk.
num_fragments_written	BIGINT	The number of written memory fragments.
num_fragments_read	BIGINT	The number of read memory fragments.
cycles_over_1_sec	BIGINT	The count of cycles that ran over 1 second for this operator instance.
cycles_over_3_sec	BIGINT	The count of cycles that ran over 3 seconds for this operator instance.
long_dispatch_queue_events	BIGINT	The count of dispatch queue events processed by the operator with a runtime greater than 0.5 seconds.
total_bin_stream_allocated_bytes	BIGINT	The total amount of memory allocated in output blocks for variable-length column data.
total_bin_stream_used_bytes	BIGINT	The total amount of memory used in output blocks for variable-length column data.
total_col_stream_allocated_bytes	BIGINT	The total amount of memory allocated in output blocks for fixed-length column data.
total_col_stream_used_bytes	BIGINT	The total amount of memory used in output blocks for fixed-length column data.
blocked_received	BIGINT	The total number of input data blocks received by this operator instance.
blocked_created	BIGINT	The total number of non-trivial output data blocks created by this operator instance.
received_block_bin_stream_unused_space_bytes	BIGINT	The total amount of wasted space across all binary streams in received data blocks (includes over-allocations as well as projected columns and dead memory). This number is NULL unless configuration parameters allow recording this statistic for each block because the calculation is expensive.
received_block_col_stream_unused_space_bytes	BIGINT	The total amount of wasted space across all column streams in received data blocks (includes over-allocations as well as projected columns and dead memory). This number is NULL unless configuration parameters allow recording this statistic for each block because the calculation is expensive.
mandatory_alloc_attempts	BIGINT	The count of cycles where the scheduler enabled mandatory allocation.
mandatory_alloc_success_count	BIGINT	The count of cycles where the scheduler enabled mandatory allocation and this operator broke an out-of-memory deadlock.
max_num_pending_blocks	BIGINT	The maximum number of pending blocks across all partitions during the lifetime of this operator.
max_num_consecutive_backup_cycles	BIGINT	The maximum number of consecutive scheduler cycles where this operator has a backup. The scheduler can use this information to help detect certain resource deadlocks.
longest_time_since_unblocked_cycle	BIGINT	The maximum time between successful attempts to run this operator from the scheduler. The scheduler can use this information to help detect certain resource deadlocks.
compiled_level	BIGINT	The level within a hierarchy of distinct operators to execute. A single plan operator can be compiled into this hierarchy.
additional_json	VARCHAR	Additional statistics for an operator instance in JSON format.
num_bloom_filters_sent	BIGINT	The number of Bloom filters the operator sent to its peers.
num_bloom_filters_received	BIGINT	The number of Bloom filters the operator received from its peers.
num_key_lists_sent	BIGINT	The number of potential index join key lists the operator sent to its peers.
num_key_lists_received	BIGINT	The number of potential index join key lists the operator received from its peers.
max_hash_table_rows	BIGINT	The maximum number of rows present in any hash table used by this operator instance.
plan_operator_estimated_output_cardinality	BIGINT	The number of rows the optimizer estimated for this plan operator to emit.
num_delayed_exceptions_generated	BIGINT	The number of rows this operator generated with delayed exception information.
num_pending_blocks_spilled	BIGINT	The number of queued pending data blocks this operator spilled to disk.
max_normal_cycle_time_ms	DOUBLE	The time, in milliseconds, of the longest normal run cycle for the operator instance.
max_oom_cycle_time_ms	DOUBLE	The time, in milliseconds, of the longest out-of-memory run cycle for the operator instance.
num_rows_received_per_plan_child	BIGINT[]	The number of rows this operator received from each of its children in the plan.
initialization_time	TIMESTAMP	The timestamp that represents when the operator began work.
num_cycles_blocked_by_temp_disk	BIGINT	The count of cycles that the operator spent waiting for input and output from the temporary disk.
num_cycles_blocked_by_external_pending_blocks	BIGINT	The count of cycles that the operator spent waiting for input and output from external pending blocks.
blobstore_callback_time_ms	DOUBLE	The time, in milliseconds, the system spends processing callbacks for temporary disk input and output for the operator instance.
num_bounding_box_lists_sent	BIGINT	The number of geospatial bounding box list filters the operator sent to its peers.
num_bounding_box_lists_received	BIGINT	The number of geospatial bounding box list filters the operator received from its peers.
target_rows_per_cycle	BIGINT	The number of rows this operator expects to process in a single cycle without exceeding time limits.
applied_dynamic_filter_stats	VARCHAR	Additional statistics about dynamic pushdown filters in JSON format.

sys.completed_queries

This table contains information about queries that finished execution in the database.

Column Name	Column Type	Column Description
query_id	UUID	A Universally Unique IDentifier (UUID) of the executed query.
user	VARCHAR	The name of the user that executed the query.
database_name	VARCHAR	The name of the database that runs this query.
database_id	UUID	The unique identifier of the database this query ran in.
sql	VARCHAR	The executed SQL statement.
sql_text_length	BIGINT	The length of the SQL statement.
referenced_tables	VARCHAR[]	The tables and views referenced by the query.
total_time	BIGINT	The total time in milliseconds the query took to run. This time includes generation_time, optimization_time, and execution_time.
generation_time	BIGINT	The time in milliseconds it took to generate the query before any processing happened.
optimization_time	BIGINT	The time in milliseconds the optimizer processed the query and built an optimized plan for the query.
execution_time	BIGINT	Time in milliseconds the query was executed by the database on the LTS nodes to return the query result.
timestamp_start	TIMESTAMP	A timestamp that represents the point in time when the query entered the system.
parsing_time	BIGINT	The time in milliseconds the query was parsing.
cache_lookup_time	BIGINT	The time in milliseconds the query spent in lookup up matching cached result sets.
validation_time	BIGINT	The time in milliseconds the query was validating.
plangen_time	BIGINT	The time in milliseconds the query plan was being generated.
queue_time	BIGINT	The time in milliseconds the query was queued in the system before any processing happened.
timestamp_optimization_start	TIMESTAMP	A timestamp that represents the point in time when optimization of the query started.
timestamp_execution_start	TIMESTAMP	A timestamp that represents the point in time when execution of the query started.
tree_probe_time	BIGINT	The time in milliseconds taken by the VM tree probe during query execution.
vm_initialization_time	BIGINT	Time in milliseconds the query was initializing on all participating nodes during execution.
timestamp_first_byte_sent	TIMESTAMP	A timestamp that represents the point in time when the first byte of the result set from the query was returned to the application.
timestamp_complete	TIMESTAMP	A timestamp that represents the point in time when the execution of the query completed from the client’s perspective.
timestamp_execution_complete	TIMESTAMP	A timestamp that represents the point in time when the internal execution of the query completed.
awaiting_client_eof_fetch_time	BIGINT	Time in milliseconds that represents the difference between when the client fetched the eof for the query and when the eof was queued internally.
rows_returned	BIGINT	The number of rows returned by the query.
bytes_returned	BIGINT	The number of bytes returned by the query.
rows_inserted	BIGINT	The number of rows that the query inserts. This value is NULL when the database does not insert any rows. The value is 0 when the database does not identify any rows to insert from a `CREATE TABLE AS SELECT` or `INSERT AS SELECT` SQL statement.
rows_deleted	BIGINT	The number of rows that the query deletes. This value is NULL when the database does not delete any rows. The value is 0 when the database does not identify any rows to delete from a `DELETE FROM TABLE` SQL statement.
update_subquery_execution_time	BIGINT	The time in milliseconds taken by the virtual machine (VM) to execute a subquery that identifies the rows for insertion or deletion. This value is NULL when you execute a SQL statement that is not one of these statements: `CREATE TABLE AS SELECT`, `INSERT AS SELECT`, or `DELETE FROM TABLE`.
transferred_bytes_per_second	BIGINT	The transfer rate in bytes per second for the result set of the query.
code	INT	The SQL code returned at the end of the query.
state	VARCHAR	The SQL state returned at the end of the query.
reason	VARCHAR	A message associated with the SQL code and SQL state.
initial_priority	DOUBLE	Priority based on the service class, session, and query-level limits.
initial_effective_priority	DOUBLE	initial_priority adjusted for total cost and memory usage.
final_effective_priority	DOUBLE	Final dynamically adjusted priority.
cost_estimate	DOUBLE	The cost estimate of the optimizer.
concurrency_service_class_name	VARCHAR	Name of the service class this query executes in.
concurrency_service_class_id	UUID	The unique identifier of the service class used for the query.
priority_adjust_factor	DOUBLE	The percentage amount by which the database adjusts the priority.
priority_adjust_time	INT	How frequently the database adjusts the priority during query execution.
cached_query	BOOLEAN	Flag that indicates whether the result was returned from the result set cache.
resultset_cached	BOOLEAN	Flag that indicates whether the result of the query was stored in the result set cache.
temp_disk_consumed	BIGINT	The approximate total temporary disk usage, in bytes, during query execution.
protocol_version	VARCHAR	The version of the client-server protocol for this connection.
client_ip	VARCHAR	The IP address of the client that executed the query.
client_name	VARCHAR	The name of the client that executed the query.
client_session_id	VARCHAR	The session id of the client that executed the query.
num_client_threads	INT	The maximum number of client threads determined by the server.
node_id	UUID	The unique identifier of the node that stores the information about this query.
participating_nodes	VARCHAR[]	The nodes that participated in the execution of this query.
ocient_db_version	VARCHAR	The Ocient® database version used to run this query.
ocient_db_commit	VARCHAR	The Ocient database git commit used to run this query.
ocient_db_dirtied_commit	VARCHAR	The dirtied Ocient database git commit used to run this query.
max_temp_disk_usage	INT	The maximum temporary disk usage for this query, expressed as a percentage.
max_elapsed_time	INT	The maximum elapsed time for this query, in seconds.
max_rows_returned	BIGINT	The maximum number of rows this query can return.
num_root_operator_instances	INT	The number of created root operator instances.
approx_system_peak_vm_node_heap_mem_bytes	BIGINT	The max observed heap memory in use during execution on a single non-SQL vm node. The memory in use is not necessarily associated with this query.
approx_system_peak_vm_cluster_heap_mem_bytes	BIGINT	The max observed heap memory in use during execution totaled across all non-SQL vm nodes. The memory in use is not necessarily associated with this query.
approx_system_peak_vm_node_huge_mem_bytes	BIGINT	The max observed huge page memory in use during execution on a single non-SQL vm node. The memory in use is not necessarily associated with this query.
approx_system_peak_vm_cluster_huge_mem_bytes	BIGINT	The max observed huge page memory in use during execution totaled across all non-SQL vm nodes. The memory in use is not necessarily associated with this query.
approx_system_peak_sql_node_heap_mem_bytes	BIGINT	The max observed heap memory in use during execution on the sql node. The memory in use is not necessarily associated with this query.
approx_system_peak_sql_node_huge_mem_bytes	BIGINT	The max observed huge page memory in use during execution on the sql node. The memory in use is not necessarily associated with this query.
driver_version	VARCHAR	The version of the client driver.
join_order_optimization_time	BIGINT	The time, in milliseconds, the join order of the query was optimized.
join_order_optimization_algorithm	VARCHAR	The algorithm used for optimizing the join order of the query.
tags	VARCHAR[]	The list of tags specified for the query.
default_schema	VARCHAR	The default schema associated with the client that executed the query.
transaction_id	UUID	The transaction scope UUID for statements executed inside an explicit transaction (with autocommit set to `OFF`). This value is NULL for autocommit statements.
transaction_prefix_len	INT	The number of high-order bits of transaction scope UUID that identify the transaction group. This value is NULL for autocommit statements.
parent_query_id	UUID	The Universally Unique IDentifier (UUID) of the parent query that spawned this child query. This value is NULL if the query is a top-level query.

sys.metric_levels

This table contains settings for metric levels.

Column Name	Column Type	Column Description
match	CHAR	Metric name match.
match_type	CHAR	The type of match for the name of the metric.
level	CHAR	Metric level.
node_id	UUID	The unique node identifier for this setting. This setting might be NULL for system-wide metrics.

sys.node_status

This table contains the status of nodes in the system and the core node monitoring metrics.

Column Name	Column Type	Column Description
node_id	UUID	The Universally Unique IDentifier (UUID) of the node.
operational_status	CHAR	The operational status of the node. Values are ACTIVE, STARTING, STOPPING, ERROR, UNKNOWN, or UNREACHABLE.
loadavg_1	DOUBLE	Average load on the node over the last 1 minute. Represents the average number of processes in the system run queue.
loadavg_5	DOUBLE	Average load on the node over the last 5 minutes. Represents the average number of processes in the system run queue.
loadavg_15	DOUBLE	Average load on the node over the last 15 minutes. Represents the average number of processes in the system run queue.
max_rss	LONG	Maximum resident set size (RSS) of the process on the node in bytes.
heap	LONG	Heap memory allocated on the node in bytes.
software_start	TIMESTAMP	Timestamp for the last start of the Ocient software on the node.
system_start	TIMESTAMP	Timestamp for the last start of the node hardware.

sys.op_inst_debug_info

This table contains information about the debug information for the operator instance.

Column Name	Column Type	Column Description
operator_id	UUID	The Universally Unique IDentifier (UUID) of the operator
node_id	UUID	The unique identifier of the node that stores the information about this operator.
silo_id	LONG	The unique identifier of the silo that stores the information about this operator.
vm_core_id	LONG	The unique identifier of the virtual machine core that stores the information about this operator.
runtime_id	INT	The process-wide unique identifier of this operator instance.
key	CHAR	The debug information (debugInfo) key.
value	CHAR	The debugInfo value.
op_inst_type	CHAR	Type of the operator Instance.
query_id	UUID	The UUID of the query that owns the operator.

sys.osn_acquisitions

This table contains information about client identifiers for Ownership Storage Number (OSN) acquisitions.

Column Name	Column Type	Column Description
node_id	UUID	The Universally Unique IDentifier (UUID) of the node.
client_id	UUID	The query identifier of the OSN acquisition.
osn	LONG	The locked OSN.
acquisition_time	TIMESTAMP	Timestamp that represents when the OSN was acquired.

sys.plans

This table contains information about the plan for each SQL statement. Not all SQL statements have a corresponding plan (e.g., some DDL statements, SQL statements that have failed parsing or optimization).

Column Name	Column Type	Column Description
query_id	UUID	The unique identifier of the SQL query that executed.
user	VARCHAR	The name of the user that executed the SQL query.
database_name	VARCHAR	The name of the database that executes this SQL query.
timestamp_start	TIMESTAMP	A timestamp that represents the point in time when the query entered the system.
plan	VARCHAR	The plan for the SQL query.

sys.queries

This table contains information about queries that the database is currently executing.

Column Name	Column Type	Column Description
query_id	UUID	A Universally Unique IDentifier (UUID) of the query that is currently running.
user	CHAR	The name of the user that executes the query.
database_name	CHAR	The name of the database that runs this query.
database_id	UUID	A unique identifier of the database this query is running in.
status	CHAR	The statuses represent how the virtual machine (VM) tracks a specific query. The statuses are: AWAITING_SLOT (The query waits for a service class slot for optimization or execution.), OPTIMIZING (The Ocient System optimizes the query. If the system cannot optimize the query, the query skips this status and changes to the QUEUED status.), QUEUED (The system initializes the query in the VM.), RUNNING (The system executes the query in the VM, either by compiling the result from the Foundation Nodes or by fetching the results from the cache.), and FINALIZING (The query is complete.).
sql	CHAR	The SQL statement that is executing.
sql_text_length	LONG	The length of the SQL statement.
referenced_tables	ARRAY(CHAR)	The tables and views referenced by the query.
total_time	LONG	The total time in milliseconds the query has run. This time includes generation_time, optimization_time, and execution_time.
generation_time	LONG	The time in milliseconds the query was generated or is still generating. This number might increase if generation is in progress.
optimization_time	LONG	The time in milliseconds the optimizer processed or is still processing the query. This number might be 0 if optimization of the query has not started yet. This number might increase if optimization is in progress.
execution_time	LONG	Time in milliseconds the query executes. This number might be 0 if execution of the query has not started yet. This number might increase if execution of the query is still in progress.
timestamp_start	TIMESTAMP	A timestamp that represents the point in time when the query entered the system.
parsing_time	LONG	Time in milliseconds the query was parsing or is still parsing. This number might be 0 if parsing of the query has not started yet. This number might increase if the query is still parsing.
cache_lookup_time	LONG	The time in milliseconds the query spent in lookup up matching cached result sets.
validation_time	LONG	Time in milliseconds the query was validating or is still being validated. This number might be 0 if validation of the query has not started yet. This number might increase if the query is still validating.
plangen_time	LONG	Time in milliseconds the query plan was being generated or is still generating. This number might be 0 if generation of the query plan has not started yet. This number might increase if the query plan is still generating.
queue_time	LONG	Time in milliseconds the query was queued or is still queuing. This number might be 0 if queuing of the query has not started yet. This number might increase if the query is still in the queue.
timestamp_optimization_start	TIMESTAMP	A timestamp that represents the point in time when optimization of the query started. This value might be 0 if optimization has not started yet.
timestamp_execution_start	TIMESTAMP	A timestamp that represents the point in time when execution of the query started. This value might be 0 if execution has not started yet.
tree_probe_time	LONG	Time in milliseconds the query was in the VM tree probe stage during execution. This number might be 0 if VM tree probe has not started yet. This number might increase if the probe is still happening.
vm_initialization_time	LONG	Time in milliseconds the query was initializing on all participating nodes during execution. This number might be 0 if the VM has not begin initializing the query for execution yet.
timestamp_first_byte_sent	TIMESTAMP	A timestamp that represents the point in time when the first byte of the result set from the query was returned to the application. This value might be 0 if execution of the query has not started yet or no bytes have been transferred yet.
rows_returned	LONG	The number of rows returned by the query. This value might be 0 if no rows have been transferred so far.
bytes_returned	LONG	The number of bytes returned by the query. This value might be 0 if no rows have been transferred so far.
rows_inserted	LONG	The number of rows that the query inserts. This value is NULL when the database does not insert any rows. The value is 0 when the database does not identify any rows to insert from a `CREATE TABLE AS SELECT` or `INSERT AS SELECT` SQL statement.
rows_deleted	LONG	The number of rows that the query deletes. This value is NULL when the database does not delete any rows. The value is 0 when the database does not identify any rows to delete from a `DELETE FROM TABLE` SQL statement.
update_subquery_execution_time	LONG	The time in milliseconds taken by the VM to execute a subquery that identifies the rows for insertion and deletion. This value is NULL when you execute a SQL statement that is not one of these statements: `CREATE TABLE AS SELECT`, `INSERT AS SELECT`, or `DELETE FROM TABLE`.
initial_priority	DOUBLE	Priority based on the service class, session, and query level limits.
initial_effective_priority	DOUBLE	initial_priority adjusted for total cost and memory usage.
effective_priority	DOUBLE	Current dynamically adjusted priority.
estimated_time	INT	The estimate of the optimizer for the time in milliseconds to process the query.
estimated_result_rows	LONG	The estimate of the optimizer for the number of rows in the result set.
estimated_result_size	LONG	The estimate of the optimizer for the size of the result set in bytes.
concurrency_service_class_name	CHAR	Name of the service class this query executes in.
concurrency_service_class_id	UUID	The unique identifier of the service class used for the query.
priority_adjust_factor	DOUBLE	The percentage amount by which the database adjusts the priority.
priority_adjust_time	INT	How frequently the database adjusts the priority during query execution.
cached_query	BOOLEAN	Flag that indicates whether the result is returned from the result set cache.
temp_disk_consumed	LONG	The approximate total temporary disk usage, in bytes, during query execution.
server_ip	IPV4	The IP address of the SQL node that executed the query.
protocol_version	CHAR	The version of the client-server protocol for this connection.
driver_version	CHAR	The version of the client driver.
client_ip	CHAR	The IP address of the client that executed the query.
client_name	CHAR	The name of the client that executed the query.
client_session_id	CHAR	The session id of the client that executed the query.
num_client_threads	INT	The maximum number of client threads determined by the server.
node_id	UUID	The unique identifier of the node that provided the information about this query.
participating_nodes	ARRAY(CHAR)	The nodes that are currently known to be participating in the execution of this query. If the query is not in the RUNNING state, this list might not include all nodes that participate.
max_temp_disk_usage	INT	The maximum temporary disk usage for this query, expressed as a percentage.
max_elapsed_time	INT	The maximum elapsed time for this query, in seconds.
max_rows_returned	LONG	The maximum number of rows this query can return.
num_root_operator_instances	INT	The number of created root operator instances.
approx_system_peak_vm_node_heap_mem_bytes	LONG	The max observed heap memory in use during execution on a single non-SQL vm node. The memory in use is not necessarily associated with this query.
approx_system_peak_vm_cluster_heap_mem_bytes	LONG	The max observed heap memory in use during execution totaled across all non-SQL vm nodes. The memory in use is not necessarily associated with this query.
approx_system_peak_vm_node_huge_mem_bytes	LONG	The max observed huge page memory in use during execution on a single non-SQL vm node. The memory in use is not necessarily associated with this query.
approx_system_peak_vm_cluster_huge_mem_bytes	LONG	The max observed huge page memory in use during execution totaled across all non-SQL vm nodes. The memory in use is not necessarily associated with this query.
approx_system_peak_sql_node_heap_mem_bytes	LONG	The max observed heap memory in use during execution on the sql node. The memory in use is not necessarily associated with this query.
approx_system_peak_sql_node_huge_mem_bytes	LONG	The max observed huge page memory in use during execution on the sql node. The memory in use is not necessarily associated with this query.
join_order_optimization_time	LONG	The time, in milliseconds, the join order of the query was optimized.
join_order_optimization_algorithm	CHAR	The algorithm used for optimizing the join order of the query.
tags	ARRAY(CHAR)	The list of tags specified for the query.
default_schema	CHAR	The default schema associated with the client that executed the query.
transaction_id	UUID	The UUID of the transaction where this statement executes. This value is NULL when the statement is not part of an explicit transaction. For `COMMIT` and `ROLLBACK` SQL statements, the value is the scope for the commit or rollback.
transaction_prefix_len	INT	The transaction scope prefix length. This value is NULL when the statement is not part of an explicit transaction.
timestamp_last_fetch	TIMESTAMP	A timestamp that represents the point in time when the application most recently fetched result data for this query. This value might be 0 if the application has not fetched any result data yet.
max_fetch_gap	LONG	The longest observed gap, in milliseconds, between consecutive result-data fetches by the application. This gap includes the still-open gap after the most recent fetch, so a large value indicates an application that has stopped consuming results.
parent_query_id	UUID	The Universally Unique IDentifier (UUID) of the parent query that spawned this child query. This value is NULL if the query is a top-level query.

sys.result_cache

This table contains queries that have cached results in the system.

Column Name	Column Type	Column Description
node_id	UUID	The Universally Unique IDentifier (UUID) of the node that holds the cache results.
query_id	UUID	The UUID of the query.
database_id	UUID	The database of the query.
service_class_id	UUID	The UUID of the service class of the query.
timestamp	LONG	The timestamp in seconds after the UNIX epoch when the result set entered the cache.
time	CHAR	The human-readable time when the result set entered the cache.
time_remaining	LONG	The amount of time remaining, in seconds, before the system purges the cached result set. This value is NULL if no finite system-level purge timeout applies, which is the case when at least one service class has defined an unlimited `cache_max_time` value.
referenced_tables	ARRAY(CHAR)	All referenced tables in this query.

sys.retention_policies

The retention_policies table shows the tables that have a retention policy enabled in the system.

Column Name	Column Type	Column Description
table_id	UUID	The Universally Unique IDentifier (UUID) of the table.
retention_interval	VARCHAR	The retention interval for the retention policy.
execution_interval	VARCHAR	The execution interval for the retention policy.

sys.sample_segment_io_pipelines

This table contains information about a sample of segment I/O processing plans used for the database table I/O during query execution.

Column Name	Column Type	Column Description
completed_time	TIMESTAMP	The time when this segment pipeline was completed.
database_name	VARCHAR	The database associated with the query to which the I/O operator belongs.
user_name	VARCHAR	The name of the user associated with the query to which the I/O operator belongs.
query_id	UUID	The Universally Unique IDentifier (UUID) of the query to which the I/O operator belongs.
operator_uuid	UUID	The UUID of the I/O operator in the query plan.
operator_id	BIGINT	An identifier of the I/O operator in the VM node that owns it.
node_id	UUID	The UUID of the node on which the I/O operator was executed.
silo_id	BIGINT	The identifier of the silo on which this I/O operator was executed.
segment_storage_id	UUID	The UUID for the segment associated with this I/O pipeline.
table_id	UUID	The UUID for the database table where this segment belongs.
segment_bounds_lower_inclusive	BIGINT	The lower bound (inclusive) on the row numbers processed in this pipeline.
segment_bounds_upper_exclusive	BIGINT	The upper bound (exclusive) on the row numbers processed in this pipeline.
pipeline	VARCHAR	A text representation of this segment I/O pipeline.

sys.scheduled_tasks

This table contains the details of scheduled tasks in the system.

Column Name	Column Type	Column Description
id	UUID	The scheduled task identifier.
admin_owner_id	UUID	The identifier of the node that is managing the scheduled task.
task_type	CHAR	The type of the task.
execution_interval_ms	LONG	The interval, in milliseconds, between task executions.

sys.schemas

This table contains the details of schemas in the system.

Column Name	Column Type	Column Description
id	UUID	The Universally Unique IDentifier (UUID) of the schema.
name	CHAR	The name of the schema.
database_id	UUID	The UUID of the database for the schema.
created_at	TIMESTAMP	The timestamp that represents when the schema was created.
updated_at	TIMESTAMP	The timestamp that represents when the schema was updated.
creator_id	UUID	The UUID of the creator for the schema.
is_implicit	BOOLEAN	Denotes the implicit creation of this schema after the creation of another object.

sys.service_role_status

This table contains the status of each service role on the system.

Column Name	Column Type	Column Description
node_id	UUID	The Universally Unique IDentifier (UUID) of the node.
service_role_id	UUID	The UUID of the service role (sys.service_roles).
service_role_name	CHAR	The human-readable name of the service role on the node.
status	CHAR	The current status of the role on this node.

sys.storage_device_metrics

This table contains monitoring metrics collected on each storage device that is controlled by Ocient.

Column Name	Column Type	Column Description
id	CHAR	Unique and persistent serial number of the drive. This serial number is not a Universally Unique IDentifier (UUID).
key	CHAR	Name of the measured statistic on the drive.
value	CHAR	Value of the named statistic on the drive.
transient	BOOLEAN	When the value is true, the value indicates that it is transient in nature and is reset when the node is restarted.
node_id	UUID	The UUID of the node that contains the storage device.
cluster_id	UUID	The UUID of the storage cluster (sys.clusters) where the node of the device belongs. This value is NULL if the node is not a member of a storage cluster.

sys.storage_device_status

This table contains real-time information about the storage devices for each online node.

Column Name	Column Type	Column Description
node_id	UUID	The Universally Unique IDentifier (UUID) of the node that contains the storage device.
device_id	INT	Identifier of the drive.
role_name	CHAR	The name of the role or roles active on the node that contains the storage device.
id	CHAR	Unique and persistent serial number of the drive. This serial number is not a UUID.
pci_address	CHAR	PCI address where the drive is mounted.
device_model	CHAR	Model information about the drive.
manufacturer	CHAR	Manufacturer information about the drive.
firmware_version	CHAR	Firmware version that is active on the drive.
capacity	LONG	Capacity, in bytes, of the drive.
utilization	LONG	Utilization, in bytes, of the drive.
endurance_percentage_used	DOUBLE	Value from 0 to 1 that represents an estimate of the percentage of the drive used, as it applies to drive endurance.
device_status	CHAR	Human-readable version of the drive status. Values are `ACTIVE`, `FAILED`, or `NOT PRESENT`.
assigned_drive_slots	ARRAY(INT)	List of drive slots assigned to this device. The list might be empty if no slots are assigned to a device.
encryption_drive_locking_supported	BOOLEAN	Indicates whether encryption drive locking (e.g., OPAL) is supported on the device.
encryption_drive_locking_enabled	BOOLEAN	Indicates whether encryption drive locking (e.g., OPAL) is enabled on the device.
encryption_drive_locking_status	CHAR	Indicates the status of drive locking (e.g., OPAL): `LOCKED` or `UNLOCKED`. If drive locking is not supported, this value is `UNLOCKED`.
silo_id	LONG	The identifier of the silo where this device is located.
cluster_id	UUID	The UUID of the storage cluster (sys.clusters) where the node of the device belongs. This value is NULL if the node is not a member of a storage cluster.

sys.subtasks

This table contains the status and details of subtasks in the system.

Column Name	Column Type	Column Description
id	UUID	The task identifier.
name	CHAR	The name of the task.
task_type	CHAR	The type of the task.
execution_type	CHAR	The type of the task execution.
location_type	CHAR	The type of the location where the task must run.
location_id	CHAR	The location where the task must run.
task_options	CHAR	The task arguments.
start_time	TIMESTAMP	The start time of the task.
end_time	TIMESTAMP	The end time of the task.
last_poll_time	TIMESTAMP	The last time that the task was polled for a status.
duration	LONG	The duration, in milliseconds, of the task.
status	CHAR	The current status of the task.
details	CHAR	The details or results of the status for the current task.
task_owner_id	UUID	The identifier of the node that is running the task.
admin_owner_id	UUID	The identifier of the node that is monitoring the task.
parent_task_id	UUID	The identifier of the parent task for the current task.
root_task_id	UUID	The identifier of the root task for the current task.
database_id	UUID	The database identifier.
parallelization	CHAR	The parallelization type of the task.
state	BINARY	The internal state of the task.
scope	UUID	The identifier of the scope for the current task.

sys.tasks

The tasks table shows the root tasks that have been started in the system.

Column Name	Column Type	Column Description
id	UUID	Task identifier
name	VARCHAR	Task name
task_type	VARCHAR	Task type
execution_type	VARCHAR	Task execution type
location_type	VARCHAR	Location type where the task must run
location_id	VARCHAR	Location where the task must run
task_options	VARCHAR	Task arguments
start_time	TIMESTAMP	Task start time
end_time	TIMESTAMP	Task end time
last_poll_time	TIMESTAMP	Last time the task was polled for status
duration	BIGINT	Task duration in milliseconds
status	VARCHAR	Current task status
details	VARCHAR	Current task status details or results
task_owner_id	UUID	Identifier of the node that is running the task
admin_owner_id	UUID	Identifier of the node that is monitoring the task
parallelization	VARCHAR	Parallelization type of the task
scope	UUID	Identifier of the scope of the task

Network

sys.channel_endpoint_parameters

This table contains details for the endpoints defined on the nodes in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the channel endpoint.
node_id	UUID	UUID of the node where the endpoint is located (sys.nodes).
name	CHAR	Name of the channel endpoint.
endpoint_type	CHAR	Type of the channel endpoint (FULL or EXTERNAL).
port	INT	Port that the channel endpoint is using.
ip_address	CHAR	IP address of the channel endpoint.
context_type	CHAR	Context type of the channel endpoint
listener_type	CHAR	Listener type of the channel endpoint
initiator_type	CHAR	Initiator type of the channel endpoint.

sys.network_interface_models

This table contains the network devices registered on the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the NIC.
manufacturer	CHAR	Manufacturer of the NIC.
model_name	CHAR	Model name of the NIC.
pci_id	CHAR	PCI identifier.
pci_subsys_id	CHAR	PCI Subsystem identifier.
driver_type	CHAR	Driver type of NIC.
bits_per_second	LONG	Bits per second.

sys.network_interface_usage_types

This table contains the uses of each network device in the system.

Column Name	Column Type	Column Description
network_interface_id	UUID	Universally Unique IDentifier (UUID) of the network interface (sys.network_interfaces).
network_interface_model_id	UUID	UUID of the network interface model (sys.network_interface_models).
network_type	CHAR	Usage type of NIC (ADMIN, DATA, EXTERNAL, LOCAL_HSI).

sys.node_network_interfaces

This table contains the network interfaces on each node.

Column Name	Column Type	Column Description
network_interface_model_id	UUID	Universally Unique IDentifier (UUID) of the network interface model (sys.network_interface_models).
network_interface_id	UUID	UUID of the network interface.
node_id	UUID	UUID of the node where this interface exists (sys.nodes).
interface_type	CHAR	Type of interface (PHYSICAL, BOND_SLAVE, BOND_MASTER).
address	CHAR	MAC address of the NIC.
name	CHAR	Name of the NIC.
netmask	CHAR	Netmask of the NIC.
gateway	CHAR	Gateway of the NIC.
pci_address	CHAR	PCI address of the NIC.
master_interface_name	CHAR	Master interface name of the NIC.
bond_type	CHAR	Network bond type of the NIC.

sys.service_role_channel_endpoints

This table contains the channel endpoint parameters that are enabled for each network type and service role.

Column Name	Column Type	Column Description
service_role_id	UUID	The Universally Unique IDentifier (UUID) of the service role of this endpoint (sys.service_roles).
channel_endpoint_parameters_id	UUID	UUID of the parameters for this endpoint (sys.channel_endpoint_parameters).
network_type	CHAR	Network type for this endpoint (ADMIN, DATA, EXTERNAL, LOCAL_HSI).

Pipelines

sys.pipeline_current_throughput

This table contains the current throughput of rows and bytes ingested (MB) by the data pipeline.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline (sys.pipelines) loading this partition.
records_throughput_per_second	DOUBLE	The throughput of the data pipeline measured as the current records loaded per second.
megabytes_throughput_per_second	DOUBLE	The throughput of the data pipeline measured as the current megabytes loaded per second.

sys.pipeline_errors

While running a pipeline, you might experience errors. These errors are often due to a mismatch between the expectation of the pipeline and the source data. System or networking errors can also occur while the pipeline is running. The sys.pipeline_errors system catalog table lists extraction, transformation, and pipeline errors.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline (sys.pipelines) that loads this partition.
extractor_task_id	UUID	UUID of the task associated with this pipeline run (sys.subtasks).
error_index	BIGINT	Error indexes start at 1 and increase monotonically as the Ocient System finds them.
error_type	VARCHAR	The type of error. Values are EXTRACTION, TRANSFORMATION, and PIPELINE_ERROR.
error_code	VARCHAR	This 5-character code identifies the type of error.
source_name	VARCHAR	For file loads, this value is the absolute path of the file that includes the S3 bucket. For Apache® Kafka® loads, the value is the topic name and partition number.
error_message	VARCHAR	The error message.
partition_id	VARCHAR	The identifier of the partition. For file loads, the value is the stream source identifier where the file is located. For Apache Kafka loads, this value is the partition number.
record_number	BIGINT	The one-based index of the processed record in the source file or Kafka partition defined in the `source_name` column. This index does not account for headers or skipped lines.
record_offset	BIGINT	The one-based offset in bytes of the record data in the source file for file loads or the Kafka offset for Kafka loads. This offset is -1 for files with non-row-based storage formats (such as Apache® Parquet™).
field_index	INT	The zero-based index of the expression in the `SELECT` clause of the `CREATE PIPELINE` SQL statement.
column_name	VARCHAR	The name of the column in the target table.
created_at	TIMESTAMP	Timestamp that represents when this error was created.

sys.pipeline_errors_historical

This view shows pipeline errors for pipelines that have been dropped from the system. The view is the historical equivalent of sys.pipeline_errors.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline that loads this partition.
extractor_task_id	UUID	UUID of the task associated with this pipeline run (sys.subtasks).
error_index	BIGINT	Error indexes start at 1 and increase monotonically as the Ocient System finds them.
error_type	VARCHAR	The type of error. Values are `EXTRACTION` (extraction), `TRANSFORMATION` (transformation), and `PIPELINE_ERROR` (pipeline error).
error_code	VARCHAR	This 5-character code identifies the type of error.
source_name	VARCHAR	For file loads, this value is the absolute path of the file that includes the Amazon® Web Services℠ (AWS℠) S3 bucket. For Kafka loads, the value is the topic name and partition number.
error_message	VARCHAR	The error message.
partition_id	VARCHAR	The identifier of the partition. For file loads, the value is the stream source identifier where the file is located. For Kafka loads, this value is the partition number.
record_number	BIGINT	The index of the processed record in the source file or Kafka partition defined in the `source_name` column. This index starts at 1. The index does not account for headers or skipped lines.
record_offset	BIGINT	The offset, in bytes, of the record data in the source file for file loads or the Kafka offset for Kafka loads. The offset starts at 1. This offset is -1 for files with non-row-based storage formats (such as Apache Parquet).
field_index	INT	The index of the expression in the `SELECT` clause of the `CREATE PIPELINE` SQL statement. This index starts at 0.
column_name	VARCHAR	The name of the column in the target table.
created_at	TIMESTAMP	Timestamp that represents when this error was created.

sys.pipeline_events

While your pipeline is running, the pipeline generates events in the sys.pipeline_events system catalog table to mark significant checkpoints when something has occurred. For details about the lists of files, see the sys.pipeline_files system catalog table. Pipeline events include messages from many different tasks. These events are the background processes that execute across different Loader Nodes during pipeline operation. For details about tasks, you can query the sys.tasks and sys.subtasks system catalog tables.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline (sys.pipelines).
task_id	UUID	UUID of the task associated with this pipeline run (sys.subtasks).
user_id	UUID	The user who initiated the event (sys.users).
event_type	VARCHAR	The type of the event. You can generate events with these types: CREATED, REPLACED, RENAMED, STARTED, STOPPING, STOPPED, QUIESCING, and QUIESCED. The Ocient System generates events with these types automatically: FILE_LISTING_STARTED, FILE_LISTING_GENERATED, FILE_LISTING_COMPLETED, COMPLETED, FAILED, EXTRACTION_STARTED, EXTRACTION_COMPLETED, EXTRACTION_FAILED, EXTRACTION_RESOURCE_EXHAUSTED, SKIPPING_PARTIALLY_LOADED_FILE, and KAFKA_PARTITION_REBALANCE.
event_message	VARCHAR	Message that provides more details about the event.
event_timestamp	TIMESTAMP	Timestamp that represents when this event occurred.

sys.pipeline_events_historical

This view shows pipeline events for pipelines that have been dropped from the system. The view is the historical equivalent of sys.pipeline_events.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline.
task_id	UUID	UUID of the task associated with this pipeline run (sys.subtasks).
user_id	UUID	The user who initiated the event (sys.users).
event_type	VARCHAR	The type of the event. You can generate events with these types: CREATED, REPLACED, RENAMED, STARTED, STOPPING, and STOPPED. The Ocient System generates events with these types automatically: FILE_LISTING_STARTED, FILE_LISTING_GENERATED, FILE_LISTING_COMPLETED, COMPLETED, FAILED, EXTRACTION_STARTED, EXTRACTION_COMPLETED, EXTRACTION_FAILED, EXTRACTION_RESOURCE_EXHAUSTED, SKIPPING_PARTIALLY_LOADED_FILE, and KAFKA_PARTITION_REBALANCE.
event_message	VARCHAR	Message that provides more details about the event.
event_timestamp	TIMESTAMP	Timestamp that represents when this event occurred.

sys.pipeline_files

For file-based loads, the sys.pipeline_files system catalog table contains one row for each file that shows the status of the file.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline (sys.pipelines).
extractor_task_id	UUID	The unique identifier of the task requested to load this file (sys.subtasks).
name	VARCHAR	Name of the file.
create_timestamp	TIMESTAMP	Timestamp that represents when this file is created.
modified_timestamp	TIMESTAMP	Timestamp that represents when this file is last modified.
size	BIGINT	Size of the file.
status	VARCHAR	The status of the file that updates throughout the process of a pipeline load. Values are: PENDING (listed file but not assigned to an extractor task), QUEUED (assigned to an extractor task but not confirmed the existence of the file), LOADING (confirmed existence of the file and began loading), LOADED (loaded data with complete success), LOADED_WITH_ERRORS (loaded data with at least one error), FAILED (failed to load), and SKIPPED (skipped loading the file).
file_index	BIGINT	Index of the file.

sys.pipeline_files_historical

This view shows pipeline files (terminal and pending) for pipelines that have been dropped from the system. The view is the historical equivalent of sys.pipeline_files.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline.
extractor_task_id	UUID	The UUID of the task requested to load this file (sys.subtasks).
name	VARCHAR	Name of the file.
create_timestamp	TIMESTAMP	Timestamp that represents when this file is created.
modified_timestamp	TIMESTAMP	Timestamp that represents when this file is last modified.
size	BIGINT	Size of the file.
status	VARCHAR	The status of the file that updates throughout the process of a pipeline load. Values are: PENDING (system listed the file but has not assigned it to an extractor task), QUEUED (system has assigned to an extractor task but has not confirmed the existence of the file), LOADING (system has confirmed the existence of the file and began loading), LOADED (system has loaded data with complete success), LOADED_WITH_ERRORS (system has loaded data with at least one error), FAILED (system has failed to load), and SKIPPED (system has skipped loading the file).
file_index	BIGINT	Index of the file.

sys.pipeline_files_pending

For file-based loads, the sys.pipeline_files system catalog table contains one row for each file that shows the status of the file.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline (sys.pipelines).
name	VARCHAR	The name of the file.
create_timestamp	TIMESTAMP	The timestamp that represents when this file is created.
modified_timestamp	TIMESTAMP	The timestamp that represents when this file is last modified.
size	BIGINT	The size of the file.
file_index	BIGINT	The index of the file.
file_transaction_id	BIGINT	The identifier of the file transaction that produced this file.

sys.pipeline_functions

This table contains all of the pipeline functions in the system.

Column Name	Column Type	Column Description
id	UUID	The Universally Unique IDentifier (UUID) of the pipeline function.
name	CHAR	The name of the pipeline function.
database_id	UUID	The UUID of the database for this pipeline function (sys.databases).
language	CHAR	The language used to write the pipeline function.
return_type	CHAR	The return type of the pipeline function.
return_type_nullable	BOOLEAN	The nullable property of the return type for the pipeline function.
definition	CHAR	The definition of the pipeline function.
creator_id	UUID	The UUID of the user or group who created the pipeline function (sys.users/sys.groups).
created_at	TIMESTAMP	The timestamp when this pipeline function was created.
altered_at	TIMESTAMP	The timestamp when this pipeline function was last updated.
argument_names	ARRAY(CHAR)	The argument names of the pipeline function, specified as an array.
argument_types	ARRAY(CHAR)	The argument types of the pipeline function, specified as an array.
argument_nullability	ARRAY(BOOLEAN)	Whether the argument types for the pipeline function are nullable, specified as an array.
imported_libraries	ARRAY(CHAR)	The import statements for the pipeline function, specified as an array.

sys.pipeline_metrics

Find metrics that measure the performance and activity of a pipeline in the sys.pipeline_metrics view. This table contains one row for each metric at a specified point in time. Metrics are either instantaneous or incremental. Instantaneous metrics reflect the current value of a counter at the time of the metric collection. Incremental metrics increase over time and reflect the cumulative value for a specified counter. The first snapshot of a metric appears when it is initialized. Afterward, the system updates the metric every 10 seconds. Each metric has a scope that defines its uniqueness. Do not aggregate metrics across different scopes. The scope columns are pipeline_id, extractor_task_id, partition_id, sink_index, file_index, work_unit_id, work_unit_attempt, and epoch_id. If the value of any of these columns is NULL, the scope applies to all values in that dimension.

Column Name	Column Type	Column Description
pipeline_id	UUID	UUID of the pipeline (sys.pipelines) loading this partition.
extractor_task_id	UUID	UUID of the task associated with this pipeline run (sys.subtasks).
partition_id	VARCHAR	The partition identifier. For file-based loads, this identifier is the stream source identifier. For Kafka loads, this identifier is the partition number.
sink_index	BIGINT	The one-based index of the table in the `CREATE PIPELINE` SQL statement or NULL for metrics that are not table-specific.
name	VARCHAR	The name of the metric.
value	BIGINT	The value of the metric.
updated_at	TIMESTAMP	Timestamp that represents when this metric was updated.
file_index	BIGINT	Index of the file.
work_unit_id	BIGINT	The work unit identifier of the unit of work dispatched to the extractor engine.
work_unit_attempt	BIGINT	The work unit attempt of the unit of work dispatched to the extractor engine.
epoch_id	BIGINT	The epoch identifier that represents when this metric was collected.

sys.pipeline_metrics_current

This table contains the most updated metrics for the data pipeline that measure the performance and activity of a pipeline in the sys.pipeline_metrics view. The scope columns are pipeline_id, extractor_task_id, partition_id, sink_index, file_index, work_unit_id, work_unit_attempt, and epoch_id. If the value of any of these columns is NULL, the scope applies to all values in that dimension.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline (sys.pipelines) loading this partition.
extractor_task_id	UUID	UUID of the task associated with this pipeline run (sys.subtasks).
partition_id	VARCHAR	The partition identifier. For file-based loads, this identifier is the stream source identifier. For Apache Kafka loads, this identifier is the partition number.
sink_index	BIGINT	The one-based index of the table in the `CREATE PIPELINE` SQL statement or NULL for metrics that are not table-specific.
name	VARCHAR	The name of the metric.
value	BIGINT	The value of the metric.
updated_at	TIMESTAMP	The timestamp that represents when this metric was updated.
file_index	BIGINT	The index of the file.
work_unit_id	BIGINT	The work unit identifier of the unit of work dispatched to the extractor engine.
work_unit_attempt	BIGINT	The work unit attempt of the unit of work dispatched to the extractor engine.
epoch_id	BIGINT	The epoch identifier that represents when this metric was collected.

sys.pipeline_metrics_historical

This view shows pipeline metrics for pipelines that have been dropped from the system. The view is the historical equivalent of sys.pipeline_metrics.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline loading this partition.
extractor_task_id	UUID	UUID of the task associated with this pipeline run (sys.subtasks).
partition_id	VARCHAR	The partition identifier. For file-based loads, this identifier is the stream source identifier. For Kafka loads, this identifier is the partition number.
sink_index	BIGINT	The index of the table in the `CREATE PIPELINE` SQL statement or NULL for metrics that are not table-specific. This index starts at 1.
name	VARCHAR	The name of the metric.
value	BIGINT	The value of the metric.
updated_at	TIMESTAMP	Timestamp that represents when this metric was updated.
file_index	BIGINT	Index of the file.
work_unit_id	BIGINT	The work unit identifier of the unit of work dispatched to the extractor engine.
work_unit_attempt	BIGINT	The work unit attempt of the unit of work dispatched to the extractor engine.
epoch_id	BIGINT	The epoch identifier that represents when this metric was collected.

sys.pipeline_metrics_info

Find the full list of available metrics in the sys.pipeline_metrics_info built-in view. This static table provides metric units, type, and descriptions.

Column Name	Column Type	Column Description
name	VARCHAR	Name of the metric.
units	VARCHAR	Units of the metric.
metric_type	VARCHAR	Type of the metric. One of INCREMENTAL or INSTANTANEOUS.
description	VARCHAR	Description of the metric.

sys.pipeline_partitions

For Apache Kafka partition-based loads, the sys.pipeline_partitions system catalog table contains one row for each partition, which contains the current offsets and record counts.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline (sys.pipelines) loading this partition.
partition_number	BIGINT	The partition number on the Kafka topic.
group_id	VARCHAR	The group identifier (group.id) of the pipeline that loads this partition.
topic_name	VARCHAR	The name of the Kafka topic where this partition belongs.
max_offset	BIGINT	The current maximum offset on this partition.
last_committed_durable_offset	BIGINT	The last committed offset on the partition that was indicated as durable by the Ocient System. This offset might lag slightly behind the actual durability of tables.
lag	BIGINT	The number of records yet to be processed in a partition, calculated as the difference between the maximum offset and the last committed durable offset.
records_processed	INT	The number of records processed (but not necessarily made durable) for this partition.
records_failed	BIGINT	The number of failed records for this partition.
updated_at	TIMESTAMP	The timestamp when the Ocient System last updated the details of this partition.

sys.pipeline_tables

This table contains all of the table identifiers currently associated with pipeline objects in the system.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline (sys.pipelines).
table_id	UUID	UUID of the table (sys.tables).
table_index	INT	The one-based index of the table in the `CREATE PIPELINE` SQL statement.

sys.pipeline_tasks

The Ocient System implements the internals of a pipeline using distributed tasks. Each pipeline has a unique pipeline identifier pipeline_id. Every time that a START PIPELINE SQL statement is executed, the Ocient System generates a run_pipeline task that serves as the main coordinator task. This task automatically partitions the source data into smaller units and creates a run_extractor task for each one. Then, the Ocient System sends each run_extractor task and executes it on a Loader Node. In the sys.pipeline_tasks system catalog table, you can see the run_pipeline and run_extractor tasks.

Column Name	Column Type	Column Description
pipeline_id	UUID	Universally Unique IDentifier (UUID) of the pipeline.
task_id	UUID	Task identifier
execution_type	VARCHAR	The type of the task.
parent_task_id	UUID	Identifier of the parent task of the task
location_type	VARCHAR	The type of location where the task must run.
location_id	VARCHAR	The identifier for the location where the task must run.
task_owner_id	UUID	Identifier of the node that is running the task
admin_owner_id	UUID	Identifier of the node that is monitoring the task
status	VARCHAR	Current task status
details	VARCHAR	The details or results of the status for the current task.
start_time	TIMESTAMP	The start time of the task.
end_time	TIMESTAMP	The end time of the task.
duration	BIGINT	Task duration in milliseconds

sys.pipelines

This table contains all of the pipelines in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the pipeline.
name	CHAR	Name of the pipeline.
database_id	UUID	UUID of the database of the pipeline (sys.databases).
loading_mode	CHAR	Specifies whether the pipeline runs in a one-time (BATCH), one-time transactional (TRANSACTIONAL), or continuous (CONTINUOUS) way.
source_type	CHAR	Source of the data (S3, KAFKA, FILESYSTEM).
data_format	CHAR	Structure of the data (CSV, DELIMITED, JSON, BINARY, PARQUET, XML, ASN1, AVRO).
created_at	TIMESTAMP	Timestamp that indicates when this pipeline was created.
altered_at	TIMESTAMP	Timestamp that indicates when this pipeline was last updated.
creator_id	UUID	The UUID of the user or group who created the pipeline (sys.users/sys.groups).
status	CHAR	Status of the pipeline (QUEUED, RUNNING, STOPPED, COMPLETED, FAILED, QUIESCING, QUIESCED).
status_message	CHAR	Status message of the pipeline.
task_id	UUID	Universally Unique IDentifier (UUID) of the task (sys.tasks).
rolehostd_version	CHAR	The rolehostd version in which this table was created
pending_files_count	LONG	The number of files with the `PENDING` status (sys.pipeline_files).
queued_files_count	LONG	The number of files with the `QUEUED` status (sys.pipeline_files).
loading_files_count	LONG	The number of files with the `LOADING` status (sys.pipeline_files).
loaded_files_count	LONG	The number of files with the `LOADED` status (sys.pipeline_files).
loaded_with_errors_files_count	LONG	The number of files with the `LOADED_WITH_ERRORS` status (sys.pipeline_files).
failed_files_count	LONG	The number of files with the `FAILED` status (sys.pipeline_files).
skipped_files_count	LONG	The number of files with the `SKIPPED` status (sys.pipeline_files).
transaction_id	UUID	The UUID of the transaction scope if this pipeline runs in a one-time transactional way.

Security

sys.security_settings

This table contains security settings.

Column Name	Column Type	Column Description
database_id	UUID	The database for this setting. This might be NULL for system-wide settings
group_id	UUID	The group for this setting when the setting is for a group
password_minimum_length	INT	Minimum password length.
password_complexity_level	INT	Password complexity requirement.
password_lifetime_days	INT	Password lifetime in days.
password_no_repeat_count	INT	Password repetition count.
password_invalid_attempt_limit	INT	Invalid password attempt limit.

Security Integrations

sys.oidc_integrations

This table enumerates the configuration for OpenID Connect Single Sign-On integrations (i.e. OKTA®). You can modify this configuration using the ALTER DATABASE <database> ALTER SSO INTEGRATION oidc DDL statement where <database> is the name of your database.

Column Name	Column Type	Column Description
security_integration_id	UUID	The id of this OpenID Connect (OIDC) Single Sign-On integration.
database_id	UUID	The id of the database using this Single Sign-On authorization to this integration.
disabled	BOOLEAN	When true, all new connections made using this Single Sign-On integration will fail.
default_group	CHAR	The group which all users of this Single Sign-On intergation are granted membership to.
issuer	CHAR	The complete URL for the OAuth 2.0 / OpenID Connect Authorization Server. This is the expected “iss” claim in an access token.
client_id	CHAR	The provider-supplied ID of this Single Sign-On integration.
public_client	BOOLEAN	When true, PKCE is used in place of a client authentication method
enable_id_token_authentication	BOOLEAN	When true, a fat ID Token is sufficient to connect to the database.
user_id_claims	ARRAY(CHAR)	The id token claim(s) used to identify users in audit trails.
additional_scopes	ARRAY(CHAR)	Request scopes for authorization requests.
additional_audiences	ARRAY(CHAR)	Additional token audience to accept when validating tokens. This is useful for Authorization Servers without a token exchange capability.
groups_claim_ids	ARRAY(CHAR)	The token claims that can be used to map the user to a Database group. You must provide `groups_claim_ids` if `groups_claim_mappings` is provided.
groups_claim_mappings	CHAR	The mappings from Provider group => Database group.
roles_claim_ids	ARRAY(CHAR)	The token claims that can be used to map the user to a Database role. You must provide `roles_claim_ids` if `roles_claim_mappings` is provided.
roles_claim_mappings	CHAR	The mappings from Provider role => Database role.
blocked_groups	ARRAY(CHAR)	Provider groups that are restricted from connecting to the database.
blocked_roles	ARRAY(CHAR)	Provider roles that are restricted from connecting to the database.
allowed_groups	ARRAY(CHAR)	Provider groups that are allowed to connect to the database.
allowed_roles	ARRAY(CHAR)	Provider roles that are allowed to connect to the database.
enable_debug_flow	BOOLEAN
redirect_host	CHAR	Specifies the redirect host used in authorization requests.
redirect_ssl	BOOLEAN	Whether to use SSL for redirect URIs.
request_scopes	ARRAY(CHAR)	All request scopes for authorization requests.
scopes_on_refresh	BOOLEAN	Whether to include the scope list in the initial refresh-token request. The default value is `false`.
allow_offline_access	BOOLEAN	Whether to include `access_type=offline` and `prompt=consent` in the authorization URL. The default value is `false`.

sys.oidc_sessions

The sys.oidc_sessions table contains information about active database connections established using an OpenID Connect (oidc) Single Sign-On integration.

Column Name	Column Type	Column Description
id	UUID	The unique identifier for this session.
user	CHAR	The Fully Qualified User Name (FQUN) of this user.
database_id	UUID	The id of the database associated with this session.
security_integration_id	UUID	The id of the OpenID Connect (OIDC) Single Sign-On integration that authorized.
has_access_token	BOOLEAN	True if the identity provider supplied an Access Token.
has_refresh_token	BOOLEAN	True if the identity provider supplied a Refresh Token.
compact_id_token	CHAR	The JWS compact serialization form of the ID Token.
id_token_json	CHAR	The ID Token claims.

sys.security_integrations

This table contains the OIDC security integrations installed on the system.

Column Name	Column Type	Column Description
id	UUID	UUID of the OIDC security integration.
database_id	UUID	UUID of the database associated with this security integration (sys.databases).
integration_type	CHAR	The type of the security integration (SSO, SCIM, LDAP, etc.).
table_descriptor	CHAR	An indirect pointer to the table where you can view the objects of this integration.
sso_integration_name	CHAR	SSO integration name, if the security integration is of type SSO
is_database_default	BOOLEAN	When true, the security integration is used as the default SSO integration for the database.

sys.sessions

This table contains information about active client connections.

Column Name	Column Type	Column Description
id	UUID	The unique identifier for this session.
user	CHAR	The Fully Qualified User Name (FQUN) of this user.
sso_protocol	CHAR	The Single Sign-On (SSO) protocol used to authorize the connection request. Null if basic password authentication was used. Tables exists for each SSO protocol using the following naming scheme: “<protocol>_sessions”. For example, all sessions created using the OpenID Connect (oidc) protocol will be listed in the “sys.oidc_sessions” table.
database_id	UUID	The id of the database associated with this session.
default_schema	CHAR	The default schema used for queries submitted by this user.
created_at_timestamp	LONG	The number of seconds from epoch the session was established.
duration	LONG	The number of seconds since the session was established.
expires_at_timestamp	LONG	The number of seconds from epoch the session expires.
refresh_expires_at_timestamp	LONG	The number of seconds from epoch the refresh token expires.
is_renewable	BOOLEAN	When the value is `true`, the database attempts to extend the session when it expires.
client_ip	CHAR	The IP address of the client.
client_name	CHAR	The human-readable name of the client.
client_session_id	CHAR	An opaque identifier provided by client implementations. Used to correlate events across application boundaries.
protocol_version	CHAR	The version of the client-server protocol for this connection.
driver_version	CHAR	The version of the client driver.
tls_enabled	BOOLEAN	True if TLS is enabled for this client connection.
node_id	UUID	The unique identifier of the node the client is connected to.
connectivity_pool_id	UUID	The unique identifier of the connectivity pool the client is connected on.
group_names	CHAR	Names of all groups in the session.
role_names	CHAR	Names of all roles in the session.
active_service_class_names	CHAR	Names of all active service classes in the session.
inactive_service_class_names	CHAR	Names of all inactive service classes in the session.

Statistics

sys.average_bb_sizes

This table contains the average size of bounding boxes in the database for geospatial data types.

Column Name	Column Type	Column Description
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
column_id	UUID	UUID of the column (sys.columns).
avg_width	DOUBLE	Average width of bounding boxes in this column.
avg_height	DOUBLE	Average height of bounding boxes in this column.

sys.average_column_sizes

This table contains the average size of each column in the database.

Column Name	Column Type	Column Description
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
column_id	UUID	UUID of the column (sys.columns).
size	DOUBLE	Average size of the column.

sys.column_cardinalities

This table contains estimates of the number of unique values in each column in the database. These estimates might not reflect recent loading or deletion activity.

Column Name	Column Type	Column Description
table_id	UUID	The Universally Unique IDentifier (UUID) of the table (sys.tables).
column_id	UUID	UUID of the column (sys.columns).
cardinality	LONG	The estimate of the number of unique values in this column.
stats_type	CHAR	Represents the type of values, either ‘COLUMN’ or ‘INNER_ARRAY’.
num_included_rows	LONG	The estimate of the total number of non-NULL rows (where this column value is not NULL) in the table that are included in this distinct estimate.
num_excluded_rows	LONG	The estimate of the total number of rows in the table that are not included in this distinct estimate.

sys.column_distributions

This table contains some statistical characteristics of the calculated Kernel Density Distribution of data in the database columns.

Column Name	Column Type	Column Description
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
column_id	UUID	UUID of the column (sys.columns).
bandwidth	DOUBLE	Smoothing parameter for the distribution.
minimum_value	DOUBLE	Minimum value found in the distribution.
maximum_value	DOUBLE	Maximum value found in the distribution.
stats_type	CHAR	Represents the type of values in the distribution, either ‘TABLE’ or ‘INNER_ARRAY’.
num_samples	LONG	The number of sampled row values included in the Kernel Density Distribution.

sys.columns_compression_info

This table contains column-level compression statistics for fixed-length columns. The system calculates statistics from the column data across all segments.

Column Name	Column Type	Column Description
node_id	UUID	Universally Unique IDentifier (UUID) of the node that owns the segments included in these statistics.
table_id	UUID	UUID of the table of this column (sys.tables).
ordinal	INT	Ordinal of the column.
is_deltadelta_enabled	BOOLEAN	Specifies whether deltadelta compression is enabled.
is_rle_enabled	BOOLEAN	Specifies whether RLE compression scheme is enabled.
is_ne_enabled	BOOLEAN	Specifies whether the NULL elimination compression scheme is enabled.
raw_size	LONG	Size of the data in bytes before compression.
compressed_size	LONG	Size of the data in bytes after compression.
num_deltadelta_blocks	LONG	Number of data blocks compressed using the deltadelta compression.
num_rle_blocks	LONG	Number of data blocks compressed using the RLE (Run-Length Encoding) compression scheme.
num_nerle_blocks	LONG	Number of data blocks compressed using a combination of the RLE and NE (NULL Elimination) compression schemes.
num_uncompressed_blocks	LONG	Number of uncompressed data blocks.
num_total_blocks	LONG	Total number of data blocks.
id	UUID	UUID of the column
name	CHAR	Name of the column

sys.segments_compression_info

This table contains segment-level compression statistics. The system calculates statistics from the data within the specified segment. All block statistics apply only to fixed-length columns within the segment. Non-block specific statistics (e.g., size information) include both fixed and variable-length columns.

Column Name	Column Type	Column Description
table_id	UUID	Universally Unique IDentifier (UUID) of the table relevant to this segment (sys.tables).
segment_id	UUID	UUID for the segment.
num_deltadelta_blocks	LONG	Number of data blocks compressed using deltadelta compression
num_rle_blocks	LONG	Number of data blocks compressed using RLE (Run-Length Encoding) compression scheme
num_nerle_blocks	LONG	Number of data blocks compressed using a combination of RLE and NE (Null Elimination) compression schemes
num_uncompressed_blocks	LONG	Number of uncompressed data blocks
num_total_blocks	LONG	Total number of data blocks
raw_size	LONG	Size of the data part before any compression and without any parity data
compressed_size	LONG	Size of the data part after compression and without any parity data + Size of the manifest part
compressed_data_part_size	LONG	Size of the data part after compression and without any parity data
num_cluster_keys	LONG	Number of unique cluster keys in the segment
num_time_buckets	LONG	Number of time buckets in the segment

sys.stats_files

This table contains information about the on-disk cache for each Foundation Node of the local table probability density functions (PDFs) and array PDFs.

Column Name	Column Type	Column Description
node_id	UUID	The Universally Unique IDentifier (UUID) of the node. References id in sys.nodes.
table_id	UUID	The UUID of the table. References id in sys.tables.
file_id	UUID	The UUID of the file.
file_type	CHAR	The type of this file. Values are ARRAY_PDF, TABLE_PDF, or TABLE_CDE.
file_size	LONG	Size in bytes of the file.
column_ordinal	LONG	Ordinal of the column where the stats file applies. The ordinal references sys.columns for the specified table. The value is NULL when file_type is TABLE_PDF.
updated_at	TIMESTAMP	Timestamp of the last update to this file.
is_stale	BOOLEAN	Whether any rows are present on disk that were not present when this stats file was last updated
rows_computed_from	LONG	Number of rows present on the node when this stats file was last updated
marked_stale_at	TIMESTAMP	Timestamp of this file being marked stale.

sys.table_cardinalities

This table contains estimates of the number of rows in the individual tables in the database. These estimates might not reflect recent loading or deletion activity.

Column Name	Column Type	Column Description
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
cardinality	LONG	Estimate of the number of rows in this table.

sys.vl_columns_compression_info

This table contains column-level compression statistics for variable-length columns. The system calculates statistics from the column data across all segments.

Column Name	Column Type	Column Description
node_id	UUID	Universally Unique IDentifier (UUID) of the node that owns the segments included in these statistics.
table_id	UUID	UUID of the table (sys.tables).
ordinal	INT	Ordinal of the column.
is_compression_enabled	BOOLEAN	Whether compression is enabled.
raw_size	LONG	Size of the data in bytes before compression.
compressed_size	LONG	Size of the data in bytes after compression.
num_lz4_rows	LONG	Number of lz4 compressed rows.
num_uncompressed_rows	LONG	Number of uncompressed rows.
num_null_rows	LONG	Number of NULL rows.
num_total_rows	LONG	Total number of rows.
id	UUID	UUID of the column
name	CHAR	Name of the column

Storage

sys.addendum_directories

This table contains all addendum directories with details about their associated directories and metadata.

Column Name	Column Type	Column Description
id	UUID	The identifier of the addendum directory within the storage cluster state.
associated_directory_group_id	CHAR	The group identifier of the associated directory.
associated_directory_ida_offset	INT	The IDA offset (the logical segment in a segment directory group) of the associated directory.
parent_segment_group_id	CHAR	The segment group identifier of the associated parent segment (sys.segment_groups).
parent_ida_offset	INT	The IDA offset of the associated parent segment.
segment_part_inventory_name	UUID	The name of the segment part inventory for the addendum part. This name references the id column of the segment_part_inventory system catalog table.
segment_part_inventory_storage_id	CHAR	The storage identifier of the segment part inventory.
num_deleted_rows	LONG	If the segment part inventory houses a delete part, this number is the number of rows subtracted from the parent segment.
is_replica_for_ida_offset	INT	The IDA offset of the parent segment that this segment part inventory replicates.

sys.current_osn

This table contains the current Ownership Sequence Number (OSN) for each cluster on the system.

Column Name	Column Type	Column Description
start_osn	CHAR	The start OSN for the cluster.
end_osn	CHAR	The end OSN for the cluster.
cluster_id	UUID	The Universally Unique IDentifier (UUID) of the cluster.

sys.data_usage

This table contains information about the amount of queryable data for each table and file type.

Column Name	Column Type	Column Description
segment_type	VARCHAR	The type of the segment.
database_name	VARCHAR	The name of the database for the table.
schema	VARCHAR	The schema of the table.
name	VARCHAR	The name of the table.
num_segments	BIGINT	The number of segments of a specific segment type.
total_rows	BIGINT	The total number of rows for segments of a specific segment type.
avg_rows_per_segment	DOUBLE	The average amount of rows per segment for a segment of a specific segment type.
total_size_gb	DOUBLE	The total size in gigabytes for all segments of a specific segment type.
avg_size_mb	DOUBLE	The average size in megabytes for segments of a specific segment type.

sys.degraded_segment_groups

This table contains information about degraded segment groups in addition to the clusters and tables where they are located.

Column Name	Column Type	Column Description
cluster_id	UUID	The unique identifier of the cluster.
cluster_name	VARCHAR	The name of the cluster.
segment_group_status	VARCHAR	The segment group status.
database_name	VARCHAR	The name of the database for the table.
schema	VARCHAR	The schema of the table.
name	VARCHAR	The name of the table.
degraded_segment_groups	BIGINT	The total number of damaged segment groups in a cluster.

sys.merge_eligibilities

This table computes and displays a count of segment groups and segment directory groups eligible for merging for each table in the Ocient System.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
database_name	VARCHAR	Name of the database for the table.
schema	VARCHAR	Name of the schema.
name	VARCHAR	Name of the table.
segment_groups_merge_eligible	BIGINT	The count of segment groups eligible for merging in the table.

sys.merge_policies

This table computes and displays the effective merge policy (enabled or disabled) for each table in the system. The system bases the policy on the default configuration and any overrides that are present for each table.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
database_name	VARCHAR	Name of the database for the table.
schema	VARCHAR	Name of the schema.
name	VARCHAR	Name of the table.
table_lts_property_string	VARCHAR	Additional properties assigned to control the Foundation role (formerly “LTS role”) behavior for the table.
effective_table_merge_policy	VARCHAR	The effective merge policy applied to this table based on the value of the global `feature_directory_merge` feature flag and an optional table merge policy override.

sys.orphaned_segments

This table contains information about orphaned segments in the cluster.

Column Name	Column Type	Column Description
storage_id	UUID	The Universally Unique IDentifier (UUID) for the orphaned segment.
owner	LONG	The identifier of the owner for this orphaned segment.
segment_type	CHAR	The type of the segment (TKT, PAGE, etc.).
node_id	UUID	The UUID of the node where the system stores the orphaned segment.
cluster_id	UUID	The UUID of the storage cluster (sys.clusters) where the node of the segment belongs. This value is NULL if the node is not a member of a storage cluster.

sys.segment_directories

This table contains all segment directories with details about parents, associated buckets, and metadata.

Column Name	Column Type	Column Description
segment_group_id	CHAR	The identifier of the segment group (sys.segment_groups).
ida_offset	INT	The unique index of the segment.
storage_id	UUID	The storage identifier of the segment directory.
owner	LONG	The owner identifier of this stored segment.
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
begin_time	LONG	Start time of the bucket of this segment directory in the time bucket column of its table.
end_time	LONG	End time of the bucket of this segment directory in the time bucket column of its table.
depth	INT	The depth of the segment group. A non-zero value indicates a segment directory group.
parent_segment_group_id	CHAR	The parent identifier of the segment directory group (sys.segment_groups).
parent_ida_offset	INT	The unique index of the parent segment, if it exists.
parent_storage_id	UUID	The storage identifier of the parent segment directory, if it exists.
parent_owner	LONG	The owner of the parent storage identifier. A zero value indicates an unowned storage identifier. Whereas a non-zero value indicates a segment owned after rebuild.
row_count	LONG	If the directory depth equals 1, then this count is the number of rows in the base leaf segment (not factoring in any DELETE SQL statements). If the depth is greater than 1, then this count is the total number of rows spanned by the intermediate directory.
created_time	TIMESTAMP	Specifies the time, in nanoseconds, when this segment group was created.
rolehostd_version	CHAR	The version of the rolehostd binary at the time of the segment group generation.

sys.segment_group_transfers

This table contains all segment group transfers between clusters.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the segment group transfer.
src_cluster_id	UUID	UUID of the source cluster (sys.clusters).
dst_cluster_id	UUID	UUID of the destination cluster (sys.clusters).
status	CHAR	The status of the transfer.
segment_group_ids	ARRAY(CHAR)	List of the identifiers of segment groups in this transfer (sys.segment_groups).
src_committed_osn	CHAR	The Ownership Storage Number (OSN) in which the segment groups are considered fully transferred from the source cluster.
dst_committed_osn	CHAR	The OSN in which the segment groups are considered fully transferred from the destination cluster.

sys.segment_groups

This table contains all segment groups stored on the cluster with details about the ownership, associated table, data structure, and other metadata.

Column Name	Column Type	Column Description
id	CHAR	The identifier of the segment group.
cluster_id	UUID	Universally Unique IDentifier (UUID) of the internal cluster of this segment group.
segment_type	CHAR	Type of the segment (TKT, PAGE, etc.).
status	CHAR	The availability and health status of the segment group.
primary_owner	UUID	For a replicated segment group, the UUID of the node that serves the segment (sys.nodes).
loader_id	UUID	The streamloader node that wrote the segment group (sys.nodes).
table_id	UUID	UUID of the table (sys.tables).
scope_id	UUID	UUID of the storage scope (sys.storage_scopes).
block_size	LONG	Disk block size used to store segments in this group in bytes.
begin_time	LONG	Inclusive minimum bucket value, specified as an integer, for all rows in this segment group. If the bucket column is a date or time value, this column has the same unit of measure as defined by the TimeKey column and you can convert or compare it to other dates or times with the appropriate conversion functions. When you specify a TIMESTAMP TimeKey, the value represents nanoseconds after the epoch.
end_time	LONG	Inclusive maximum bucket value, specified as an integer, for all rows in this segment group. If the bucket column is a date or time value, this column has the same unit of measure as defined by the TimeKey column and you can convert or compare it to other dates or times with the appropriate conversion functions. When you specify a TIMESTAMP TimeKey, the value represents nanoseconds after the epoch.
begin_timestamp	TIMESTAMP	Inclusive minimum bucket value, specified as a TIMESTAMP, for all rows in this segment group. If the bucket column is a unitless integer, this column is NULL.
end_timestamp	TIMESTAMP	Inclusive maximum bucket value, specified as an integer, for all rows in this segment group. If the bucket column is a unitless integer, this column is NULL.
coding_algorithm	CHAR	Coding algorithm used to erasure code this group (NO_CODING, PQ_PARITY, REED_SOLOMON).
coding_block_size	INT	The unit size that is erasure coded in bytes.
coding_threshold	INT	The number of coding blocks required to rebuild all blocks in a coding line.
coding_width	INT	The number of coding blocks in a coding line.
replication	INT	The number of replicas of each segment in the group.
parity_cycle	INT	Parity cycle that the system calculates by multiplying by the coding_width to return the number of segments in the segment group.
created_time	TIMESTAMP	Specifies the time, in nanoseconds, when this segment group was created.
rolehostd_version	CHAR	The version of the rolehostd binary at the time of the segment group generation.
commit_hash	CHAR	The commit hash of the rolehostd binary at the time of the segment group generation.
timestamp	CHAR	The build timestamp or commit timestamp of the rolehostd binary at the time of the segment group generation.
build_user	CHAR	The build user of the rolehostd binary at the time of the segment group generation.
depth	INT	The depth of the segment directory group. A non-zero value indicates a segment directory group.
removal_type	CHAR	The method by which a segment group was removed.
visibility	CHAR	The visibility of the segment group.
regeneration_group	UUID	The UUID of the regeneration group.
regeneration_types	ARRAY(CHAR)	Type of regeneration with these values: `REINDEX` for reindexing and `REAGG` for reaggregation.

sys.segment_part_inventory

This table contains information about the inventory of markers for segments. For example, a marker denotes whether the database can delete a specific segment part. One segment can have multiple inventories, whereas each inventory contains only one marker.

Column Name	Column Type	Column Description
id	UUID	The universally unique identifier (UUID) of the inventory that stores the markers for segments.
ida_offset	INT	The IDA offset (the logical part of the segment) denotes the position within the segment.
parent_segment_group_id	CHAR	The identifier of the segment group where the parent segment is located.
parent_ida_offset	INT	The IDA offset where the parent segment is located within the segment group.
num_deleted_rows	LONG	The number of deleted rows.
operation_id	UUID	The UUID of the query where the Ocient System creates the inventory (the `DELETE` SQL statement). The system shares this UUID across all inventories created in the same operation.
subsuming_type	CHAR	The type of the marker within the inventory.
node_id	UUID	The UUID of the node that contains the latest segment with the part.

sys.segment_part_redundancy_info

This table contains the redundancy strategy for each table and the segment part type.

Column Name	Column Type	Column Description
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
part_name	CHAR	Name of the segment part.
redundancy_type	CHAR	The redundancy strategy for this part (COPY, PARITY).

sys.segment_parts

This table contains the segment parts associated with each segment group. These segment parts belong to queryable segments. The table does not display parts that belong to quarantined segments or segments not yet activated.

Column Name	Column Type	Column Description
segment_group_id	CHAR	The identifier of the segment group (sys.segment_groups).
ida_offset	INT	Unique index of the segment. The original copy of this segment part belongs to this segment.
name	CHAR	Name of the segment part.
part_type	CHAR	Partial identifier of a segment part (DATA, INDEX, MANIFEST, STATS).
size	LONG	Size of the segment part in bytes.
segment_ida_offset	INT	Unique index of a segment within a segment group.
segment_lba_offset	LONG	Offset of this segment part within the Logical Block Address (LBA) on disk.

sys.segments

This table contains the individual segments stored in each storage cluster including basic information about the segment and the segment group where the segment belongs. The table contains queryable segments only. Quarantined segments or segments not yet activated do not appear in this table (they are present in the sys.stored_segments table).

Column Name	Column Type	Column Description
segment_group_id	CHAR	The identifier of the segment group (sys.segment_groups).
segment_type	CHAR	Type of the segment (TKT or PAGE).
ida_offset	INT	Unique index of the segment in the segment group.
row_count	LONG	Count of accessible rows in the segment. If the count is unknown, the value is NULL.
storage_id	UUID	Universally Unique IDentifier (UUID) for the stored segment (sys.stored_segments.storage_id).
table_id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
segment_size	LONG	Size of the segment in bytes.
root_segment_directory_group_id	CHAR	The identifier of the root-level segment directory group to which this segment belongs. For root-level directories and non-subsumed segments, this value is the same as the `segment_group_id` column.
root_segment_directory_group_ida_offset	INT	The unique index of the directory in the root-level directory group to which this segment belongs.
begin_time	LONG	Inclusive minimum bucket value, specified as an integer, for all rows in this segment. If the bucket column is a date or time value, this column has the same unit of measure as defined by the TimeKey column and you can convert or compare it to other dates or times with the appropriate conversion functions. When you specify a TIMESTAMP TimeKey, the value represents nanoseconds after the epoch.
end_time	LONG	Inclusive maximum bucket value, specified as an integer, for all rows in this segment. If the bucket column is a date or time value, this column has the same unit of measure as defined by the TimeKey column and you can convert or compare it to other dates or times with the appropriate conversion functions. When you specify a TIMESTAMP TimeKey, the value represents nanoseconds after the epoch.
begin_timestamp	TIMESTAMP	Inclusive minimum bucket value, specified as a TIMESTAMP, for all rows in this segment. If the bucket column is a unitless integer, this column is NULL.
end_timestamp	TIMESTAMP	Inclusive maximum bucket value, specified as an integer, for all rows in this segment. If the bucket column is a unitless integer, this column is NULL.
owner	LONG	Owner identifier of this stored segment.
num_deleted_rows	LONG	Number of rows deleted from this segment.
page_size_with_replication	LONG	Specifies the replication of stored pages.

sys.stop_osn_reap_requests

This table contains the list of active requests that guard the Ownership Sequence Number (OSN) from reaping. Each request guards the OSN that was current at creation (guard_osn): the cluster might reap older OSNs, but never ones newer than the guard. Guarding ends for an OSN after the system removes all requests guarding it.

Column Name	Column Type	Column Description
request_id	UUID	The Universally Unique IDentifier (UUID) of the request.
created_at	TIMESTAMP	The timestamp that represents when the request was created.
description	CHAR	The description of the request.
cluster_id	UUID	The UUID of the cluster.
guard_osn	LONG	The OSN this request guards. Reaping might advance the cluster start OSN up to this value but never past it.

sys.storage_capacity

This table contains information about storage capacity for each node.

Column Name	Column Type	Column Description
node_id	UUID	Universally Unique IDentifier (UUID) of the node (sys.nodes).
capacity_bytes	LONG	Approximate storage capacity of the node in bytes.
cluster_id	UUID	UUID of the cluster (sys.clusters).

sys.storage_device_files

This table contains the list of all files for each storage device.

Column Name	Column Type	Column Description
node_id	UUID	The Universally Unique IDentifier (UUID) of the node.
device_id	INT	The numeric identifier of the drive.
storage_id	UUID	The UUID of the file.
owner	LONG	Owner of the storage identifier. The zero value indicates an unowned storage identifier. A non-zero value indicates a segment owned after the rebuild.
segment_type	CHAR	The type of the segment (TKT, PAGE, etc.).
parent_id	UUID	The UUID of the parent file (for addendum parts).
normalized_drive_slot	INT	Theoretical drive slot, if applicable, that was computed from the segment group identifier. The Ocient System expects to allocate the file to this slot.
abnormal_placement	BOOLEAN	Specifies whether the Ocient System places the segment abnormally. If this value is `true`, the segment is not located on the drive slot that was computed from the segment group identifier.
segment_group_id	CHAR	The identifier of the segment group (sys.segment_groups).
cluster_id	UUID	The UUID of the storage cluster (sys.clusters) where the node of the device belongs. This value is NULL if the node is not a member of a storage cluster.

sys.storage_scopes

This table contains the storage scopes defined in the system with the number of rows, page groups, and segment groups in each storage scope.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the storage scope.
row_count	LONG	The number of rows that have been currently loaded onto the scope.
num_page_groups	LONG	The number of page groups currently present in the scope.
num_tkt_segment_groups	LONG	Number of TKT segment groups currently present in the scope.
scope_prefix_len	INT	The prefix length of the scope.
transaction_timeout	LONG	The transaction timeout of the scope.
creation_timestamp	TIMESTAMP	The timestamp when the scope was created.
last_activity_timestamp	TIMESTAMP	The timestamp when the scope was last updated.
cluster_id	UUID	UUID of the cluster (sys.clusters).

sys.storage_spaces

This table contains information for all defined storage spaces.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the storage space.
name	CHAR	Name of the storage space.
is_system_storage_space	BOOLEAN	Is this storage space the system storage space
block_size	INT	Unit size that is erasure coded in bytes.
total_width	INT	The coding width. (The N in an M of N parity configuration.)
parity_width	INT	Within a coding line, the number of blocks dedicated to storing parity information.
parity_type	CHAR	The methodology used to compute parity blocks (P+Q, XOR, REED SOLOMON, REPLICATION, NONE).
parity_cycles	INT	Parity cycles calculated by multiplying by total_width to return the number of segments in the segment group.
page_replication	INT	The total number of page replicas, which includes the original page.

sys.storage_used

This table contains information about storage utilization for each node.

Column Name	Column Type	Column Description
node_id	UUID	Universally Unique IDentifier (UUID) of the node (sys.nodes).
table_id	UUID	UUID of the table (sys.tables).
used_bytes	LONG	Approximate storage utilization of the table on the node in bytes.
cluster_id	UUID	UUID of the cluster (sys.clusters).

sys.stored_leaf_segment_part_inventory

This table contains the physical location for the inventory of markers for all segments.

Column Name	Column Type	Column Description
segment_group_id	CHAR	The identifier of the segment group (sys.segment_groups).
cluster_id	UUID	Universally Unique IDentifier (UUID) of the cluster (sys.clusters).
ida_offset	INT	Index of the segment in the segment group.
storage_id	UUID	UUID of the stored segment.
owner	LONG	Owner identifier of the stored segment.
root_directory_group_id	CHAR	Root segment directory identifier, if it has been subsumed.
node_id	UUID	UUID of the node (sys.nodes).
start_osn	CHAR	The starting Ownership Sequence Number (OSN) for the latest OSN range of the inventory of markers.
end_osn	CHAR	The ending OSN for the latest OSN range of the inventory of markers.
kind	CHAR	The kind of segment (DISK, VIRTUAL).
visibility	CHAR	The visibility of the inventory of markers.
toc_name	UUID	UUID of the inventory of markers. A NULL UUID denotes the main inventory of markers of the segment.
row_count	LONG	The number of rows in the base leaf segment (not factoring in any DELETE SQL statements).

sys.stored_segments

This table contains details for each segment that is saved to disk.

Column Name	Column Type	Column Description
segment_group_id	CHAR	The identifier of the segment group (sys.segment_groups).
cluster_id	UUID	Universally Unique IDentifier (UUID) of the cluster (sys.clusters).
ida_offset	INT	Index of the segment in the segment group.
storage_id	UUID	UUID of this stored segment.
owner	LONG	Owner identifier of this stored segment.
node_id	UUID	UUID of the node (sys.nodes).
status	CHAR	The current status of the segment (MISSING, INTACT, etc.).
kind	CHAR	The kind of segment (DISK, VIRTUAL).
start_osn	CHAR	The start OSN for the latest OSN range of the segment.
end_osn	CHAR	The end OSN for the latest OSN range of the segment.
visibility	CHAR	The visibility of the stored segment. Valid values are VISIBLE (segment is queryable), QUARANTINED (segment is not queryable), HIDDEN (segment is hidden), and INVALID (segment is invalid). During segment generation, the Ocient System can make segments hidden or invalid for a short period before the segments become queryable.
abnormal_placement	BOOLEAN	Whether the segment is placed abnormally.
row_count	LONG	Count of rows in the segment.
segment_size	LONG	Size of the segment in bytes.

sys.unhealthy_segments

This table contains information about nodes with damaged, missing, or invalid segments.

Column Name	Column Type	Column Description
node_id	UUID	The unique identifier of the node.
node_name	VARCHAR	The name of the node.
status	VARCHAR	The status of the stored segment. Values include INTACT, MISSING, and DAMAGED.
num_unhealthy_segments	BIGINT	The total number of unhealthy segments in the node.

System

sys.clusters

This table contains all clusters defined in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the cluster (sys.clusters.id).
name	CHAR	Name of the cluster.
cluster_type	CHAR	Type of the cluster.
storage_space_ids	ARRAY(UUID)	UUIDs of the storage spaces used in this cluster (sys.storage_spaces.id).

sys.compute_configurations

This table contains the compute configurations for nodes and clusters in the system.

Column Name	Column Type	Column Description
cluster_id	UUID	Universally Unique IDentifier (UUID) of the cluster that this node belongs to (sys.clusters.id).
node_id	UUID	UUID of the node (sys.nodes.id).
level	LONG	Execution level for this node or cluster.
csn	LONG	Compute sequence number.

sys.locks

This table contains information of the currently queued and granted lock requests.

Column Name	Column Type	Column Description
request_id	UUID	The unique identifier of the lock.
owner_identifier	CHAR	The unique identifier of the owner of the lock. Should be of format `<node uuid with lowercase chars>.<locking system>.<locking reason>.<process uuid>`
lock_scope_id	CHAR	An identifier on the scope of the lock, should be of format `<system>.<type>.<target unique identifier>`
lock_type	CHAR	Whether the lock is to READ or WRITE on the elements in its scope.
status	CHAR	The status of the lock, QUEUED means that it is waiting to be accepted, GRANTED means it is currently active.
create_time	LONG	The creation time, in milliseconds, of the lock relative to the underlying RAFT Consensus Log. This value is not a UNIX timestamp.
last_refresh_time	LONG	The time, in milliseconds, when the lock was last refreshed relative to the underlying RAFT Consensus Log. This value is not a UNIX timestamp
priority_id	UUID	The identifier for lock prioritization.
create_timestamp	TIMESTAMP	The timestamp that represents when the lock was created on the Raft leader.
last_refresh_timestamp	TIMESTAMP	The timestamp that represents when the lock was most recently refreshed on the Raft leader.

sys.lts_cluster_info

This table contains the storage clusters defined in the system.

Column Name	Column Type	Column Description
cluster_id	UUID	Universally Unique IDentifier (UUID) of the storage cluster (sys.clusters.id).
storage_space_ids	ARRAY(UUID)	UUIDs of the storage spaces used in this cluster (sys.storage_spaces.id).

sys.node_clusters

This table contains nodes that are members of each cluster.

Column Name	Column Type	Column Description
cluster_id	UUID	Universally Unique IDentifier (UUID) of the cluster that this node belongs to.
node_id	UUID	UUID of the node.
ordinal	INT	Ordinal of this node.

sys.nodes

This table contains all nodes defined in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the node.
unique_node_number	INT	Ordinal number assigned to the node.
name	CHAR	Name of the node.
status	CHAR	Status of the node (ACCEPTED, UNACCEPTED, INVALID).
hostname	CHAR	Hostname of the node.
software_version	CHAR	Version of the software running on this node.
kernel_version	CHAR	Version of the kernel running on this node.
system_version	CHAR	Version of Linux® running on this node.
system_memory	LONG	Amount of available system memory in bytes.
sockets	INT	Number of CPU sockets.
cores_per_socket	INT	Number of cores per CPU socket.
hugepages_1gb	INT	Number of configured 1GB huge pages.
hugepages_2mb	INT	Number of configured 2MB huge pages.
hyperthreaded	BOOLEAN	Whether the node is hyperthreaded.

sys.service_roles

This table contains the service roles defined on each node.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the service role.
service_role_type	CHAR	Type of the service role.
node_id	UUID	UUID of the node. (sys.nodes.id)
level	CHAR	Execution level of the node.

System Information

sys.function_signatures

This table contains all function signatures defined in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the function signature.
name	CHAR	Name of the function.
return_type	CHAR	Return type of the function.
arg_types	ARRAY(CHAR)	The types of all arguments in the function.
function_type	CHAR	Type of the function.

sys.reserved_words

This table contains all reserved words defined in Ocient.

Column Name	Column Type	Column Description
reserved_word	CHAR	A word that is a reserved keyword for use by Ocient.

sys.sql_messages

This table contains information about the SQL errors and warnings that the Ocient System can produce.

Column Name	Column Type	Column Description
name	CHAR	SQL error or warning name.
code	INT	SQL error or warning code.
state	CHAR	SQL error or warning state.
reason	CHAR	Description of the SQL error or warning.
is_error	BOOLEAN	Whether or not this SQL message is an error or warning.

sys.system_information

This table contains information about the Ocient System.

Column Name	Column Type	Column Description
id	UUID	Universally Unique Identifier (UUID) for the overall system.
name	CHAR	The name of the system. This system assigns this name randomly using a UUID by default. You can change the name using the `ALTER SYSTEM RENAME` SQL statement.
current_compatible_software_version	INT	The minimum software version for the system compatibility.
allowed_compatible_software_version	INT	The software version that prevents the use of features in the newest version of the system to keep backward compatibility with the specified version in the `current_compatible_software_version` column.
feature_directory_merge	CHAR	The value of the `directory merge` feature flag for segment directory groups.

sys.system_table_columns

This table contains all columns for tables available in the system catalog.

Column Name	Column Type	Column Description
id	UUID	The Universally Unique IDentifier (UUID) of this column (sys.columns).
name	CHAR	The name of this column.
data_type	CHAR	The data type of the column.
nullable	BOOLEAN	Whether or not this column is nullable.
ordinal	LONG	The ordinal of this column.
schema_name	CHAR	The name of the schema to which this column belongs.
table_name	CHAR	The name of the table to which this column belongs.
description	CHAR	Description of the column.

sys.system_table_constraints

This table contains all constraints for tables available in the system catalog.

Column Name	Column Type	Column Description
constraint_name	CHAR	The name of this constraint.
constraint_type	CHAR	The type of the constraint with valid values: `PRIMARY_KEY` or `FOREIGN_KEY`.
schema_name	CHAR	The name of the schema to which this constraint belongs.
table_name	CHAR	The name of the table to which this constraint belongs.
column_names	ARRAY(CHAR)	The names of the columns where the constraint applies.
referenced_schema_name	CHAR	The schema of the referenced table. This value is NULL for non-foreign key constraints.
referenced_table_name	CHAR	The name of the referenced table. This value is NULL for non-foreign key constraints.
referenced_column_names	ARRAY(CHAR)	The names of the columns where the referenced constraint applies. This value is NULL for non-foreign key constraints.

sys.system_tables

This table contains all tables available in the system catalog.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the table (sys.tables).
schema	CHAR	Schema of the table.
name	CHAR	Name of the table.
table_type	CHAR	Type of the table.
category	CHAR	Category of the table.
description	CHAR	Description of the table.

User Management

sys.group_roles

This table contains roles that belong to each group.

Column Name	Column Type	Column Description
group_id	UUID	Universally Unique IDentifier (UUID) of the group (sys.groups).
role_id	UUID	UUID of the role (sys.roles).

sys.groups

This table contains information, including the service class, related to each group.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the group.
name	CHAR	Name of the group.
database_id	UUID	UUID of the database (sys.databases).
service_class_id	UUID	UUID of the service class of this group (sys.service_classes).

sys.privileges

This table contains information related to the privileges granted in Ocient.

Column Name	Column Type	Column Description
timestamp	TIMESTAMP	Timestamp of the grant given.
grantor	CHAR	The user who granted this privilege.
grantee	CHAR	The user who received this privilege.
privilege	CHAR	The privilege for the grant.
privilege_target	CHAR	The type of object to which the privilege applies. This value is NULL if the privilege applies to the granted object.
object_type	CHAR	The type of object on which this privilege was granted.
object_id	UUID	Universally Unique IDentifier (UUID) of the object.
grantable	BOOLEAN	Whether the user can grant this privilege to another user.

sys.rights

This table contains information related to the rights granted in Ocient.

Column Name	Column Type	Column Description
entity_type	CHAR	Owner of the right (user, group, role).
entity_id	UUID	Universally Unique IDentifier (UUID) of the owner.
target_type	CHAR	Type of the target.
target_id	UUID	UUID of the target.
value	CHAR	The right being given (CREATE, READ, UPDATE, DELETE, SECURITY).

sys.roles

This table contains roles in the system.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the role.
database_id	UUID	UUID of the database of this role (sys.databases).
name	CHAR	Name of this role.
description	CHAR	Detailed description of this role.

sys.user_groups

This table contains the users that belong to groups in Ocient.

Column Name	Column Type	Column Description
user_id	UUID	Universally Unique IDentifier (UUID) of the user (sys.users).
group_id	UUID	UUID of the group where the user belongs (sys.groups).

sys.user_roles

This table contains roles associated with each user.

Column Name	Column Type	Column Description
user_id	UUID	Universally Unique IDentifier (UUID) of the user (sys.users).
role_id	UUID	UUID of the role (sys.roles).

sys.users

This table contains information about users in Ocient.

Column Name	Column Type	Column Description
id	UUID	Universally Unique IDentifier (UUID) of the user.
user_name	CHAR	Username.
database_id	UUID	UUID of the database.
first_name	CHAR	First name of the user.
last_name	CHAR	Last name of the user.
email	CHAR	Email address of the user.
password_updated_at	TIMESTAMP	Timestamp that represents the date and time for the last time the password for this user was updated
state	CHAR	State.
invalid_login_attempts	INT	Number of invalid login attempts after the last successful login.
password_days_remaining	INT	Number of days remaining until the password expires. If you do not configure the password expiration, this value is NULL.

Workload Management

sys.service_classes

This table shows information about the service classes used for workload management in Ocient. For more details on workload management, see the corresponding section of the user documentation.

Column Name	Column Type	Column Description
id	UUID	The ID of the service class.
database_id	UUID	The ID of the database that has the service class.
name	CHAR	The name of the service class.
max_temp_disk_usage	LONG	The maximum temporary disk usage for a query running with this service class.
max_elapsed_time	LONG	The maximum time, in seconds, that a query running with this service class can run.
max_concurrent_queries	LONG	The maximum number of queries that can run concurrently with this service class.
max_rows_returned	LONG	The maximum number of rows a query running with this service class can return.
max_query_memory_bytes	LONG	The per-VM-node memory limit, in bytes, for a query running with this service class. -1 means unlimited.
max_query_cpu_seconds	LONG	The per-VM-node CPU-time limit, in seconds, for a query running with this service class. -1 means unlimited.
scheduling_priority	DOUBLE	The initial priority for a query running with this service class.
cache_max_bytes	LONG	The maximum number of bytes in a result set, if it is eligible for cache storage, for a query running with this service class.
cache_max_time	LONG	The maximum amount of time, in seconds, that the system caches rows for a query running with this service class.
max_elapsed_time_for_caching	LONG	The maximum amount of time, in seconds, that a query running with this service class can run and have its result set cached. This value must be higher than the `max_elapsed_time` value.
max_columns_in_result_set	LONG	The maximum number of columns allowed in the result set of a query running with this service class.
priority_adjustment_factor	DOUBLE	The amount that the system adjusts the query priority in each time interval, as specified by the `priority_adjustment_time` value, for a query running with this service class.
priority_adjustment_time	LONG	The frequency, in seconds, of each priority adjustment for a query running with this service class.
min_priority	DOUBLE	The minimum value of the adjusted priority for a query running with this service class.
max_priority	DOUBLE	The maximum value of the adjusted priority for a query running with this service class.
statement_text	CHAR	Specifies the comparison string to use for matching statement text to queries. This string takes the format specified by the statement_text_matcher_type field.
statement_text_matcher_type	CHAR	Specifies the type of string comparison to use for matching statement text to queries. This value must be LIKE or REGEX and is required if you specify a ‘statement_text’ field.
half_parallelism	BOOLEAN	Determines whether the VM uses half of the available cores for each operator in this query rather than the maximum possible parallelism. Setting this value can reduce overhead and latency for queries, but it significantly decreases data throughput. So, updating this value should be rare outside of Ocient Support work.
load_balance_shuffle	BOOLEAN	Determines whether the optimizer includes a load-balancing network operator above I/O for each query. This field can add latency to queries, but it will likely increase data throughput. If unset, the choice is left to the optimizer. So, updating this value should be rare outside of Ocient Support work.
parallelism	INT	Determines the number of parallel cores to use for each operator in the VM. This value must be nonzero. If unset, the choice is left to the VM. So, updating this value should be rare outside of Ocient Support work.
minimize_query_debug_records	BOOLEAN	This value determines whether the optimizer and VM should attempt to limit debug metrics and logging for queries associated with this service class. Limiting query observability may improve performance for workloads running many queries per second.
memory_optimal_strategy	BOOLEAN	This value determines whether the optimizer and VM should attempt to lower the memory requirements of a query regardless of potential execution time penalties. So, updating this value should be rare outside of Ocient Support work.
is_system_managed	BOOLEAN	This value indicates whether or not the service class is a system managed service class.

sys.service_classes_for_user

The service_classes_for_user table shows groups that each user belongs to and the service class for each group.

Column Name	Column Type	Column Description
user_name	VARCHAR	Username.
group_name	VARCHAR	Name of the group.
service_class_name	VARCHAR	The name of the service class.

Information Schema

Last modified on September 23, 2026

​Alphabetical List of Tables

​Alphabetical List of Categories

​Catalog Table Details

​Configuration

​sys.config

​sys.node_config

​Connectivity Pools

​sys.connectivity_pool_participants

​sys.connectivity_pools

​Databases

​sys.columns

​sys.constraints

​sys.databases

​sys.function_parameters

​sys.functions

​sys.global_map_table_info

​sys.index_advisor_enabled

​sys.index_columns

​sys.index_recommendations

​sys.index_usage

​sys.indexes

​sys.locations

​sys.procedure_parameters

​sys.procedures

​sys.tables

​sys.tkt_table_clustering_column_indices

​sys.tkt_table_info

​sys.user_mappings

​sys.view_columns

​sys.views

​Index Advisor

​sys.index_advisor_column_usages

​sys.index_advisor_index_usages

​Loading

​sys.load_errors

​sys.load_events

​sys.load_metrics

​sys.segment_batch_events

​sys.segment_generation_events

​Machine Learning

​sys.association_rules_models

​sys.bagging_models

​sys.boosting_models

​sys.dbscan_models

​sys.decision_tree_models

​sys.feature_selection_models

​sys.feature_selection_steps

​sys.feedforward_network_models

​sys.gaussian_discriminant_analysis_models

​sys.gaussian_mixture_models

​sys.gradient_boosted_trees_models

​sys.k_means_models

​sys.k_nearest_neighbors_models

​sys.linear_combination_regression_models

​sys.linear_discriminant_analysis_models

​sys.logistic_regression_models

​sys.machine_learning_model_options

​sys.machine_learning_models

​sys.mlp_models

​sys.multiple_linear_regression_models

​sys.naive_bayes_models

​sys.nonlinear_regression_models

​sys.polynomial_regression_models

​sys.principal_component_analysis_models

​sys.random_forest_models

​sys.regression_tree_models

​sys.simple_linear_regression_models

​sys.stacking_models

​sys.support_vector_machine_models

​sys.survival_regression_models

​sys.undersampling_models

​sys.vector_autoregression_models

​Model Context Protocol

​sys.mcp_tools

​Monitoring

​sys.active_operator_instances

​sys.active_query_scheduler_info

​sys.all_operator_instances

​sys.completed_operator_instances

​sys.completed_queries