Resolved Onyx upgrade issue health monitor

[DeltaFox2211]

Dear all

After i upgraded to plesk Onyx i have encounterd a problem with the health monitor.

I get the following Error

{"status":"error","statusMessages":[{"status":"error","class":"","content":"rrdtool execution failed: ERROR: No DS called 'value' in '\/usr\/local\/psa\/var\/health\/data\/localhost\/df\/df-root.rrd'

Actually i have no idea how and why this error appears maybe i someone knows how to deal with this

Thanks

 

[IgorG]

Try to rebuild this file /usr/local/psa/var/health/data/localhost/df/df-root.rrd with following steps:

# /etc/init.d/sw-collectd stop
# mv /usr/local/psa/var/health/data/localhost/df/df-root.rrd /root/
# /etc/init.d/sw-collectd start

If it not help, try to reinstall Health Monitor with

# plesk installer --select-release-current --remove-component health-monitor
# plesk installer --select-release-current --install-component health-monitor

 

[websavers]

Thanks Igor!

We had to basically do that with all the files in the /df/ folder by running:

# /etc/init.d/sw-collectd stop
# tar -cvzf /root/df_backup.tgz /usr/local/psa/var/health/data/localhost/df/df-root.rrd/*
# rm -rf /usr/local/psa/var/health/data/localhost/df/df-root.rrd/*
# /etc/init.d/sw-collectd start

That fixed it up.

 

[websavers]

Turns out I was wrong about this. After removing those files, disk monitoring is simply disabled entirely -- there's no "Disk" option showing in the Plesk Server Health Monitor any longer.

I tried restarting Plesk services and reinstalling the health monitor like this:

# plesk installer --select-release-current --remove-component health-monitor
# plesk installer --select-release-current --install-component health-monitor

I tried backing up and removing the entire 'localhost' folder, which seemed like it worked in that upon restart of sw-collectd the localhost folder was recreated, however then not only was missing Disk but Network too! Only Services, Memory, and CPU were left. And still no disks! So I restored the old config.

Then I noticed the `df` command was throwing errors about named / bind mounts and found this thread: https://talk.plesk.com/threads/solved-problem-health-monitor-server.336118/ which says to unmount the bad named mounts like this:

# umount /var/named/chroot/etc/named.rfc1912.zones
# umount /var/named/chroot/etc/named.iscdlv.key
# umount /var/named/chroot/etc/named.root.key

I did so, then used the "Detect Hardware Changes" button and now everything is back!