Excessive I/O on database server after the "Off-Hours Database Maintenance" ran (4245646)

Excessive I/O on database server after the "Off-Hours Database Maintenance" ran

After the "Off-Hours Database Maintenance" job ran, the external MySQL database is producing heavy I/O on the database server. Only restarting the Foglight server stops the I/O.

The server is purging expired data. This is the processing that runs, when a user goes to the "Data Management" dashboard and selects the "Purge" button for a topology object. This is very expensive processing. It essentially processes every table in the database and discards the data for the objects that have been purged. As such the purge processing should really only be performed if there is a significant amount of data that is to be purged.

To identify if there are a lot of purge processes, please run the script below via "Administration | Tooling | Script Console":

return "Purge operations: " + server.PersistenceService.getAllScheduledObservationPurges()

The following script can be run to cancel all scheduled purged:

server.PersistenceService.getAllScheduledObservationPurges().each{
    server.PersistenceService.cancelObservationPurge(it);
}
return "Purge operations cancelled."