[xymon] Monitoring /tmp usage

[Rob Munsch]

Ah gotcha.  Silly as it is, to include tmpfs but leave the rest of the EXCLUDE intact, I take the current

cat /proc/filesystems | grep nodev | awk '{print $2}' | xargs echo | sed -e 's! ! -x !g'

and turn it into 

cat /proc/filesystems | grep nodev | grep -v tmpfs | awk '{print $2}' | xargs echo | sed -e 's! ! -x !g'

and I seem to get the result I want.  Thanks all :)

> -----Original Message-----
> From: Henrik "Størner [mailto:henrik at hswn.dk]
> Sent: Thursday, January 20, 2011 7:23 AM
> To: xymon at xymon.com
> Subject: Re: [xymon] Monitoring /tmp usage
> 
> In <4D3751E5.6050705 at umdnj.edu> Ryan Novosielski <novosirj at umdnj.edu>
> writes:
> 
> >On 01/19/2011 03:38 PM, Rob Munsch wrote:
> >> [df]
> >> Filesystem         1024-blocks      Used Available Capacity Mounted on
> >> /dev/hdv1            3768053780 1050316584 2717737196      28% /
> >>
> >> [mount]
> >> /dev/hdv1 on / type ufs (defaults)
> >> none on /proc type proc (defaults)
> >> none on /tmp type tmpfs (size=128m,mode=1777,nosuid,noexec,nodev)
> >
> >It is possible that your problem stems from the fact that /tmp is type
> >"none" and that a grep looking for /dev would fail. I'm not exactly sure
> >what Xymon looks for, but I had to manually change some scripts
> >associated with BB to look for even ZFS filesystems which do not have
> >devices under / (or at least they don't appear to from df's POV).
> 
> Precisely. The "df" data does not have info about /tmp, so you need
> to look at the Xymon client script running on the server to have
> it report data for non-physical filesystems. Look at the
> ~xymon/client/bin/xymonclient-OSNAME.sh script (or "hobbit" ditto).
> By default, non-physical filesystems are excluded.
> 
> Regards,
> Henrik
> 
> 
> To unsubscribe from the xymon list, send an e-mail to
> xymon-unsubscribe at xymon.com
>