Matterhorn will query its own
job table every minute or so, and since older jobs are not removed from this table it will eventually reach a point where these queries become a noticeable drain on system resources. To clean out the table:
- Locate your matterhorn install and check etc/config.properties for the database you are using (default is /usr/local/matterhorn/etc/config.properties)
- Open up that database via your browser of choice and select the `mh_jobs` job` table
- Delete rows where the operation value is NOT START_WORKFLOW (we want to retain those jobs for later reference). You can use
date_createdto scope this delete via time.
- Your db table should be much smaller now.
delete from mh_job where operation <> 'START_WORKFLOW' and date_created < '2018-12-01';