Today we are introducing a new switch for all checks: ‑‑alarm_limit=0|1|2

When would we need this new switch? It will most probably be used for Raidstatus or SnapMirrorState, every time when an administrator thinks that an alarm marked as CRITICAL is too far fetched and WARNING would be enough.

 

An example:

$ ./check_netapp_pro SnapMirrorState ‑H toaster
NETAPP_PRO SNAPMIRRORSTATE CRITICAL - 21 snapmirror relations checked, 6 critical and 0 warning
isp-nfs1_dest:isp_nfs1_ki : broken-off (CRITICAL)
...

In order to set this alarm as WARNING instead of CRITICAL, simply use this setting ‑‑alarm_limit=1.

For our example:

$ ./check_netapp_pro SnapMirrorState ‑H toaster ‑‑alarm_limit=1
NETAPP_PRO SNAPMIRRORSTATE WARNING - 21 snapmirror relations checked, 0 critical and 21 warning
isp-nfs1_dest:isp_nfs1_ki : broken-off (WARNING)
...

Note: --alarm_limit can also be set to 0 or 2. The first option sets the alarm to OK and the second would have the same effect, as if the option is no set at all. This is because UNKNOWN (3) conditions will not be corrected by --alarm_limit and always remain UNKOWN.