Check ALL partitions/disks with one check - feature request

[ludmilmm]

It would be nice to be able to use one check only with NCPA to check all partitions (i.e. /, /boot, etc.) on Linux or disks in Windows (i.e. C:, D:, etc.). Currently, you need to set up separate checks. With check_disk and NSClient++ with check_nrpe, you can do this easily:

Windows

/usr/local/nagios/libexec/check_nrpe -H 192.168.x.x -c alias_disk
OK All 5 drive(s) are ok|'C:\ used'=349.18101GB;362.36562;407.66132;0;452.95702 'C:\ used %'=77%;79;89;0;100 'D:\ used'=83.0071GB;186.30312;209.59101;0;232.8789 'D:\ used %'=35%;79;89;0;100 'E:\ used'=0B;0;0;0;0 'F:\ used'=0B;0;0;0;0 '\\?\Volume{4d36d4c6-efab-11e0-a8e7-806e6f6e6963}\ used'=7.38542GB;10.20937;11.48554;0;12.76171 '\\?\Volume{4d36d4c6-efab-11e0-a8e7-806e6f6e6963}\ used %'=57%;79;89;0;100

Linux

/usr/local/nagios/libexec/check_nrpe -H 192.168.x.x -c check_disk -a '-w 80 -c 90'
DISK OK - free space: / 5013 MB (30% inode=78%); /dev/shm 1962 MB (100% inode=99%); /boot 347 MB (77% inode=99%);| /=11638MB;17464;17454;0;17544 /dev/shm=0MB;1882;1872;0;1962 /boot=103MB;396;386;0;476

[ccztux]

This would be very nice!

[jm-welch]

This is a feature I've been wanting for a while, but keep forgetting to ask for. What I've got right now is the NRPE check_disk plugin in the ./plugins folder, with a second script called check_all_disks that's just got the command line for the check_disk script to check everything I care about all at once.

What makes it useful to me (and therefore, what I'd love to see possible in NCPA):

• Can check multiple disks using the same or different thresholds for each
• Option to only include output from disks that do not pass the check in the longserviceoutput (so if everything's OK, you get perfdata, but only a single serviceoutput line of "Disks OK"

[varnar]

I dont think it is OK to have all in one.
My concern is related to graphs.
For example you checking for /, /home, /opt, /mnt, /share
You getting perf data for all of them.
Now, you removing the "/share" mount point and you starting to be in trouble.
You adding new one to existing - same, you are in trouble.
Your RRD file are expecting to have 5, but you provide them 4 or 6.

All in one can be for system resources monitoring, which is not changing :)

[ccztux]

This doesnt happen, if you store every ds in a single file, like you could do with pnp4nagios:
http://docs.pnp4nagios.org/pnp-0.6/rrd_convert
I think it is better to have less checks than more.

[varnar]

The plugin will be used not only at Nagios Core.
People are using it at NagiosXI as well.

[jomann09]

After looking into this a bit more, it may actually be possible though would require some big changes to some nodes. I'll look into it for a future version but will probably not be in 2.0.0.

[jvandermeulen]

Thank you for your update Jake.

[guglez]

I'm also interested in this feature.

[jomann09]

No longer have a milestone for this. The good news is this will start working once we update psutils to the next version, when it is released! Closing this because of that :)

[jomann09]

This was closed because of a bug fix, this feature request hasn't been implemented but will be in 3.0.0

[sawolf]

Updating this since the same thing came up from #972 - I do think this is a reasonable feature request, and I think it should be tractable to our new NCPA developers if they agree. That said, I don't think it'll be coming in 3.0.0.

[dxb74]

I would like to see this as well. NRPE currently let's you check against ALL mounted/lettered drives and lets you filter out ones you don't care about. Would like to see that same functionality.