Symptom
The system seems slow.
By default, the jobs.aggregates.maintainer.frequency parameter is set to 1 hour. This means that the aggregate maintainer, stats checker, and orphan checker run every hour, and the aggregate maintainer/stats checker executes the following small descriptive queries:
- 'USE as_adventure'
- 'SHOW TABLES'
- simple count(*) on aggregate tables
Depending on the number of aggregates, although the queries are small, the cumulative effect might impact resources which could result in a slow-down of "heart-beat" performance queries.
In those cases, the jobs.aggregates.maintainer.frequency parameter can be set to 3 hours, or as needed, to reduce the number of orphan/stat checks for aggregates. It is a hidden parameter that can be edited through UI.
Recommendation: Please contact the AtScale Support team prior to changing this parameter.
Procedure
Here are the steps to edit the jobs.aggregates.maintainer.frequency parameter:
Step 1. Navigate to Settings --> Engine --> SHOW CUSTOM SETTINGS
Step 2. Add Setting Name "jobs.aggregates.maintainer.frequency" and Setting value of "3 hours".
Step 3. Save
In order for the changes to take effect, the engine needs to be restarted.
Step 4. Perform a status check
$ /opt/atscale/current/bin/atscale_service_control status
Step 5. Restart the engine
$ /opt/atscale/current/bin/atscale_service_control restart engine
Step 6. Do another status check to make sure the engine pid has changed
$ /opt/atscale/current/bin/atscale_service_control status
Note: /opt/atscale/ is the default AtScale software home, please change it to match your environment setup.
If you have opened an AtScale Support case, as recommended, then please download the summary zip from the Design Center UI and upload it to the AtScale FTP site so that Support can confirm that the configuration change is effective.
Reference: