nihlaeth / nagios_check_smartmon Goto Github PK

Right now, root privileges are required. This is a problem, because I intend to use it with check_by_ssh. The user I assigned for this has limited privileges for obvious reasons. I thought to fix it by simply setting the setuid bit. Unfortunately (though with good reason) - this does not work on scripts, only on binaries.

cython --embed would be a possible solution, though I think it might be safer to provide a binary wrapper for smartctl. Needs further thought.

I've had disks fail with no indication save the error count in the smart log. This should be checked for!

Psutil reports partitions in use, and from there we extract the physical devices. If you use raid, encryption or lfs, the mapped devices are reported, and we can't get the physical devices from that through psutil.

Either we need to ls/grep from the dev directory for the actual physisal devices, which is bound to be buggy, or we accept the limitations of the -a/--all-disks option.

When I switched to subprocess, I assumed that a non-zero exit status would always mean some error in accessing the device or executing the command. However, when there's something wrong with the disk, smartctl indicates this in the return value as well.

Option:
We do it the easy way, we ignore the exit status and live with the fact that the parser will complain if command execution fails.

Option:
We interpret the exit status and only issue an error message & skip the parser if the command failed.

Colleagues,

plugin is hardcoded to expect "PASSED" in following code snippet

# check health status
if health_status != "PASSED":
return_status = 2
device_status += "CRITICAL: device does not pass health status "

Although in some cases smartctl reports "SMART Health Status: OK"

root@server:~# smartctl -H /dev/sdd
smartctl 6.2 2013-07-26 r3841 [x86_64-linux-3.13.0-74-generic] (local build)
Copyright (C) 2002-13, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF READ SMART DATA SECTION ===
SMART Health Status: OK

which leads to CRITICAL status.

Reproducible: always