View the query log

The query log records queries executed on your {{% product-name %}} cluster. Use it to monitor query performance, find slow-running queries, and troubleshoot failed executions.

[!Note]

Query logging is not enabled by default

The query log is disabled by default on all clusters because it generates additional ingest and storage overhead and is intended primarily for troubleshooting, not continuous monitoring. To enable it for your cluster, contact InfluxData support.

Use the Admin UI or the influxctl query command to view the query log.

{{< tabs-wrapper >}} {{% tabs %}} Admin UI influxctl {{% /tabs %}} {{% tab-content %}} {{< admin-ui-access >}}

3. Open the cluster you want to inspect and go to Query History.

If query logging is enabled for your cluster, any admin user can access the query log in the Admin UI automatically; no database token is required.

In Query History you can:

• Search by Database Token ID to see queries run with a specific token.
• Filter by:
  • Status (for example, success, failure)
  • Database
  • Query type (for example, SQL, InfluxQL)
  • Source (for example, User Queries, UI)
  • Time range (for example, last 24 hours)

The table lists each query with its status, query text, database, query type, duration, and timestamp. You can use the column headers to sort (for example by duration or time).

{{< img-hd src="/img/influxdb3/cloud-dedicated-admin-ui-query-log-list-view.png" alt="Query History list view in the Admin UI with search, filters, and table" />}}

You can also expand a row to see more details about that execution.

{{< img-hd src="/img/influxdb3/cloud-dedicated-admin-ui-query-log-detail-view.png" alt="Query History detail view in the Admin UI" />}}

{{% /tab-content %}} {{% tab-content %}}

Use the influxctl query command to run SQL against the _internal database and query_log table. Query log entries are stored in the _internal database.

1. If you haven't already, download and install the influxctl CLI, and then configure an influxctl connection profile for your cluster.
2. Create a database token that has read access to the _internal database. Replace {{% code-placeholder-key %}}DATABASE_TOKEN{{% /code-placeholder-key %}} in the examples below with your {{% token-link "database" %}}.
3. Run the query subcommand with --database and --language (and optionally --config). Global flags such as --config must come before the command; query flags such as --database, --language, and --token must come after query.

Examples

List recent successful queries with compute duration above a threshold (for example, 0.6 ms):

influxctl query \
  --token DATABASE_TOKEN \
  --database _internal \
  --language sql \
  'SELECT * FROM query_log WHERE success = '\''true'\'' AND compute_duration_ns > 600000 LIMIT 10'

Filter by namespace (database) and time range:

influxctl query \
  --token DATABASE_TOKEN \
  --database _internal \
  --language sql \
  'SELECT * FROM query_log WHERE namespace_name = '\''my_database'\'' AND time >= now() - INTERVAL '\''1 day'\'' LIMIT 50'

Example output:

| auth_id        | compute_duration_ns | phase   | query_type | query_text                                              | success | time                    |
|----------------|---------------------|---------|------------|---------------------------------------------------------|---------|--------------------------|
| token-id-xxxx  |             2314333 | success | sql        | SELECT * FROM query_log WHERE success = 'true' AND ...  | true    | 2026-02-25T00:30:30Z     |
| token-id-yyyy  |          3673637621 | success | sql        | SELECT * FROM my_measurement WHERE time > now() - ...  | true    | 2026-02-25T00:28:57Z     |
| token-id-yyyy  |          1443145654 | success | sql        | SELECT COUNT(*) FROM query_log WHERE ...                | true    | 2026-02-25T00:29:02Z     |
+----------------+---------------------+---------+------------+---------------------------------------------------------+---------+--------------------------+

{{% /tab-content %}} {{< /tabs-wrapper >}}

Query log data and columns

The query_log table in _internal includes the following columns:

[!Note]

Use string literals for status columns

In the query_log table, success and running are stored as strings ('true' or 'false'), not booleans. In SQL predicates, use string comparison—for example, success = 'true' or running = 'false'—to avoid type coercion errors.