When a CloudLinux user hits LVE limits, appropriate faults are generated and lvestats package generates Server processes snapshot.
Snapshot is a list of running applications and a list of running MySQL queries right after the faults happened.
Snapshots allow users to investigate the reason for account hitting its limits.
Several snapshots are generated for each incident.
An incident is a state when faults are generated in a close time period.
The time period is configurable.
By default, if faults are generated in 300 seconds time period, and these are considered as a single incident.
First, check the log and status of the service.
$ sudo tail -n 50 /var/log/lve-stats.log
$ sudo systemctl status lvestats
Second, check and update (if needed) to the latest stable kernel "$version" and reboot the server.
This is best done during a period where the server has minimal activity.
Another possible reason is a large DB.
So backup and recreate it using the following commands:
$ sudo systemctl stop lvestats
$ sudo tar -zcvf /root/lveinfo_backup_$(date +%Y-%m-%d).tar.gz /var/lve/
$ sudo mv /var/lve/lvestats2.db{,.old}
$ sudo lve-create-db --recreate
$ sudo systemctl start lvestats
The snapshot configuration options are available at:
/etc/sysconfig/lvestats.config/SnapshotSaver.cfg
period_between_incidents = 300, by default, time in seconds snapshots_per_minute = 2, by default, the maximum number of snapshots per minute max_snapshots_per_incident = 10, by default, maximum number of snapshots for an incident
