The CANCEL QUERY statement cancels a running SQL query.
Considerations
- Schema changes are treated differently than other SQL queries. You can use
SHOW JOBSto monitor the progress of schema changes andCANCEL JOBto cancel schema changes that are taking longer than expected. - In rare cases where a query is close to completion when a cancellation request is issued, the query may run to completion.
Required privileges
Members of the admin role (include root, which belongs to admin by default) can cancel any currently active queries. User that are not members of the admin role can cancel only their own currently active queries. To view and cancel another non-admin user's query, the user must be a member of the admin role or must have the VIEWACTIVITY and CANCELQUERY parameters set.
Synopsis
Parameters
| Parameter | Description |
|---|---|
query_id |
A scalar expression that produces the ID of the query to cancel.CANCEL QUERY accepts a single query ID. If a subquery is used and returns multiple IDs, the CANCEL QUERY statement will fail. To cancel multiple queries, use CANCEL QUERIES. |
select_stmt |
A selection query whose result you want to cancel. |
Response
When a query is successfully cancelled, CockroachDB sends a query execution canceled error to the client that issued the query.
- If the canceled query was a single, stand-alone statement, no further action is required by the client.
- If the canceled query was part of a larger, multi-statement transaction, the client should then issue a
ROLLBACKstatement.
Examples
Cancel a query via the query ID
In this example, we use the SHOW QUERIES statement to get the ID of a query and then pass the ID into the CANCEL QUERY statement:
> SHOW QUERIES;
+----------------------------------+---------+----------+----------------------------------+----------------------------------+--------------------+------------------+-------------+-----------+
| query_id | node_id | username | start | query | client_address | application_name | distributed | phase |
+----------------------------------+---------+----------+----------------------------------+----------------------------------+--------------------+------------------+-------------+-----------+
| 14dacc1f9a781e3d0000000000000001 | 2 | mroach | 2017-08-10 14:08:22.878113+00:00 | SELECT * FROM test.kv ORDER BY k | 192.168.0.72:56194 | test_app | false | executing |
+----------------------------------+---------+----------+----------------------------------+----------------------------------+--------------------+------------------+-------------+-----------+
| 14dacc206c47a9690000000000000002 | 2 | root | 2017-08-14 19:11:05.309119+00:00 | SHOW CLUSTER QUERIES | 127.0.0.1:50921 | | NULL | preparing |
+----------------------------------+---------+----------+----------------------------------+----------------------------------+--------------------+------------------+-------------+-----------+
> CANCEL QUERY '14dacc1f9a781e3d0000000000000001';
Cancel a query via a subquery
In this example, we nest a SELECT clause that retrieves the ID of a query inside the CANCEL QUERY statement:
> CANCEL QUERY (SELECT query_id FROM [SHOW CLUSTER QUERIES]
WHERE client_address = '192.168.0.72:56194'
AND username = 'mroach'
AND query = 'SELECT * FROM test.kv ORDER BY k');
CANCEL QUERY accepts a single query ID. If a subquery is used and returns multiple IDs, the CANCEL QUERY statement will fail. To cancel multiple queries, use CANCEL QUERIES.See also
Was this helpful?
