Hi, On Sat, Mar 29, 2014 at 09:13:08AM +0100, Daniel Pocock wrote:
> I'm using check_raid for an md RAID (managed by mdadm) > > However, I get errors if mptctl is not loaded: > > # /usr/lib/nagios/plugins/check_raid > open /dev/mptctl: No such file or directory > Try: mknod /dev/mptctl c 10 220 > Make sure mptctl is loaded into the kernel > OK: md:[md0(raid1):UU] > > > The errors are on stderr, regular status info is on stdout > > Loading mptctl manually makes the errors go away: > > # modprobe mptctl > > but it would be nice if I didn't have to make any workaround like this. The "proper" fix would be not having "mpt-status" installed :) One could also add "--autoload" to the mpt-status call, but I am not sure this is a good idea. > If I don't have the mpt-statusd package on my system at all, check_raid > gives spurious warnings: > > # /usr/lib/nagios/plugins/check_raid > OK: md:[md0(raid1):UU]mpt:mpt-status program not found This is fixed upstream since 3.0 and in Debian since 5.20130307. > Apart from this spurious output, it does actually work correctly for my > md RAID and correctly detected a disk failure this week. > > The system in question does have MPT hardware, but the hardware RAID > feature is not in use. Mhh... -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
