Zabbix Agent should report about RAM ECC-Errors

[thooge]

Hello,

i had some days ago a serious server breakdown due to RAM failure.
So i checked where to look up for ECC errors to monitor them.

Seems so the zabbix agend has no key for that?

I know i can realize that with an UserParameter, but this is important
enough to be implemented directly inside the agent.

The corrected and uncorrected errors should be visible in
/sys/devices/system/edac/mc/mc0/ce_count
and
/sys/devices/system/edac/mc/mc0/ue_count

In my opinion a value > 0 inside ue_count should be so seriously
to shutdown the server immediately and get repaired.

Kind regards
Thomas

[mgielissen]

I created a template: https://github.com/mgielissen/zabbix...inux_edac.yaml