[Nagios-devel] check_disk behaviour for SAN hiccups
Posted: Wed Jun 18, 2003 9:03 am
A firmware upgrade was applied to our Hitachi SAN last night. One of =
the hosts attached to it burped and didn't recover. Here's one of the =
alerts that Nagios gave us:
[quote]
***** Nagios 1.1 *****
Notification Type: PROBLEM
Service: Space on /ora_backup
Host: itdmln16
Address: itdmln16
State: WARNING
Date/Time: Tue Jun 17 22:52:19 CDT 2003
Additional Info:
Disk /ora_backup not mounted or nonexistant
[end quote]
Should this only be a 'warning'?
When one of my peers tried to do an 'ls' on the root directory, he got =
the brief listing of the various files, directories and mount points. =
When he tried to do an "ls -l", he got a couple I/O errors and then the =
long listing for all the files, directories and mount points, minus the =
2 SAN-based mount points.
Granted, this is a funky result for a single system; all the other hosts =
connected to the same SAN didn't have any such problem. Still, it would =
have been nice if the plugin had returned a 'critical' instead of a =
'warning', as that notification would have also been sent to the on-call =
pager.
jc
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]
the hosts attached to it burped and didn't recover. Here's one of the =
alerts that Nagios gave us:
[quote]
***** Nagios 1.1 *****
Notification Type: PROBLEM
Service: Space on /ora_backup
Host: itdmln16
Address: itdmln16
State: WARNING
Date/Time: Tue Jun 17 22:52:19 CDT 2003
Additional Info:
Disk /ora_backup not mounted or nonexistant
[end quote]
Should this only be a 'warning'?
When one of my peers tried to do an 'ls' on the root directory, he got =
the brief listing of the various files, directories and mount points. =
When he tried to do an "ls -l", he got a couple I/O errors and then the =
long listing for all the files, directories and mount points, minus the =
2 SAN-based mount points.
Granted, this is a funky result for a single system; all the other hosts =
connected to the same SAN didn't have any such problem. Still, it would =
have been nice if the plugin had returned a 'critical' instead of a =
'warning', as that notification would have also been sent to the on-call =
pager.
jc
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]