Query history system table reference

Preview

This system table is in Public Preview. To access the table, the schema must be enabled in your system catalog. For more information, see Enable system table schemas.

This article includes information on the query history system table, including an outline of the table’s schema.

Important

To access the query history system table, you must enable the query schema. For instructions on enabling system schemas, see Enable system table schemas.

Using the query history table

The query history table, located at system.query.history, includes records for every SQL statement run using SQL warehouses. The table includes account-wide records from all workspaces in the same region from which you access the table.

By default, only admins have access to the system table. If you would like to share the table’s data with a user or group, Databricks recommends creating a dynamic view for each user or group. See Create a dynamic view.

Query history system table schema

The query history system table uses the following schema:

Column name	Data type	Description	Example
account_id	string	ID of the account.	11e22ba4-87b9-4cc2 -9770-d10b894b7118
workspace_id	string	The ID of the workspace where the query was run.	1234567890123456
statement_id	string	The ID that uniquely identifies the execution of the statement. You can use this ID to find the statement execution in the Query History UI.	7a99b43c-b46c-432b -b0a7-814217701909
session_id	string	The Spark session ID.	01234567-cr06-a2mp -t0nd-a14ecfb5a9c2
execution_status	string	The statement termination state. Possible values are: FINISHED: execution was successful FAILED: execution failed with the reason for failure described in the accompanying error message CANCELED: execution was canceled	FINISHED
compute	struct	A struct that represents the type of compute resource used to run the statement and the ID of the resource where applicable. The type value will be either WAREHOUSE or SERVERLESS_COMPUTE.	{ type: WAREHOUSE, cluster_id: NULL, warehouse_id: ec58ee3772e8d305 }
executed_by_user_id	string	The ID of the user who ran the statement.	2967555311742259
executed_by	string	The email address or username of the user who ran the statement.	example@databricks.com
statement_text	string	Text of the SQL statement. If you have configured customer-managed keys, statement_text is empty.	SELECT 1
statement_type	string	The statement type. For example: ALTER, COPY, and`INSERT`.	SELECT
error_message	string	Message describing the error condition. If you have configured customer-managed keys, error_message is empty.	[INSUFFICIENT_PERMISSIONS] Insufficient privileges: User does not have permission SELECT on table 'default.nyctaxi_trips'.
client_application	string	Client application that ran the statement. For example: Databricks SQL, Tableau, and Power BI.	Databricks SQL
client_driver	string	The connector used to connect to Databricks to run the statement. For example: Databricks SQL Driver for Go, Databricks ODBC Driver, Databricks JDBC Driver.	Databricks JDBC Driver
total_duration_ms	bigint	Total execution time of the statement in milliseconds ( excluding result fetch time ).	1
waiting_for_compute_duration_ms	bigint	Time spent waiting for compute resources to be provisioned in milliseconds.	1
waiting_at_capacity_duration_ms	bigint	Time spent waiting in queue for available compute capacity in milliseconds.	1
execution_duration_ms	bigint	Time spent executing the statement in milliseconds.	1
compilation_duration_ms	bigint	Time spent loading metadata and optimizing the statement in milliseconds.	1
total_task_duration_ms	bigint	The sum of all task durations in milliseconds. This time represents the combined time it took to run the query across all cores of all nodes. It can be significantly longer than the wall-clock duration if multiple tasks are executed in parallel. It can be shorter than the wall-clock duration if tasks wait for available nodes.	1
result_fetch_duration_ms	bigint	Time spent, in milliseconds, fetching the statement results after the execution finished.	1
start_time	timestamp	The time when Databricks received the request. Timezone information is recorded at the end of the value with +00:00 representing UTC.	2022-12-05T00:00:00.000+0000
end_time	timestamp	The time the statement execution ended, including results fetch time. Timezone information is recorded at the end of the value with +00:00 representing UTC.	2022-12-05T00:00:00.000+00:00
update_time	timestamp	The time the statement last received a progress update. Timezone information is recorded at the end of the value with +00:00 representing UTC.	2022-12-05T00:00:00.000+00:00
read_partitions	bigint	The number of partitions read after pruning.	1
pruned_files	bigint	The number of pruned files.	1
read_files	bigint	The number of files read after pruning.	1
read_rows	bigint	Total number of rows read by the statement.	1
produced_rows	bigint	Total number of rows returned by the statement.	1
read_bytes	bigint	Total size of data read by the statement in bytes.	1
read_io_cache_percent	int	The percentage of bytes of persistent data read from the IO cache.	50
from_results_cache	boolean	TRUE indicates that the statement result was fetched from the cache.	TRUE
spilled_local_bytes	bigint	Size of data, in bytes, temporarily written to disk while executing the statement.	1
written_bytes	bigint	The size in bytes of persistent data written to cloud object storage.	1
shuffle_read_bytes	bigint	The total amount of data in bytes sent over the network.	1

View the query profile for a record

To navigate to a query’s query profile based on a record in the query history table, do the following:

1. Identify the record of interest, then copy the record’s statement_id.

2. Reference the record’s workspace_id to ensure you are logged in to the same workspace as the record.

3. Click Query History in the workspace sidebar.

4. In the Statement ID field, paste the statement_id on the record.

5. Click the name of a query. An overview of query metrics appears.

6. Click See query profile.