MRTG problems with RRD returning "got: ???/???"

[msbensonstk]

Yes, I do, and from the interface from that RRD file actually. It's last data point was around Feb. 28 @ 9:58PM

[tgriep]

So there isn't any data from the 28th of February to today for that interface?

Can you go to this folder and find that interface and do the rrdtool dump and attach the file?

Code: Select all

/usr/local/nagios/share/perfdata/

[msbensonstk]

The RRD dump is 5MB and the max upload size is 2. Is there another way I can get it to you? It has several years of data in it.

[tgriep]

That is all we have for file uploads. Can you zip it?

[msbensonstk]

Sorry, didn't even cross my mind to zip it. See attached.

[tgriep]

Can you go to that service, click on the Advanced Tab.
Are you seeing Performance Data in the details?
Can you upload a screen capture of that window so I can look at it?

Could you go to the following WIKI link and go through it to see if that fixes the problems?
http://support.nagios.com/wiki/index.ph ... h_Problems

[msbensonstk]

There is perf data coming through, however it's a value of 0 (see attached screen shot), which I'm assuming is due to the RRD database being filled with NaN -- when the RRD check goes to the RRD database to scrape the 5mn traffic delta it just gets 0.

The graph it's self is working fine since the perf data is returning 0, it's graphing 0.

[tgriep]

Have you gone through the WIKI link from my earlier post?

Run these and post back the results.

Code: Select all

tail -50 /usr/local/nagios/var/perfdata.log | grep TIMEOUT
tail -50 /usr/local/nagios/var/npcd.log | grep "MAX load reached"
LANG=C LC_ALL=C /usr/bin/mrtg /etc/mrtg/mrtg.cfg
ll /var/lib/mrtg/
service crond status
cat /etc/cron.d/mrtg

Are any of the checks returning valid data?

I don't think the rrd files can be repaired so you may have to delete them.

[msbensonstk]

Yes, I ran through the wiki steps with no changes.

The output is in the attached file as there was too much to paste in here without making things hard to read.

All non-MRTG RRD database checks are returning valid results. All MRTG RRD database checks are returning 0.

The perfdata databases are the one's with the actual badwidth that Nagios is graphing, so if we need to delete the MRTG RRD's that should be fine, but we need the historic perfdata Nagios has collected from the MRTG RRDs.

[tgriep]

These errors that you had need to be fixed.

2015-03-04 15:07:31: ERROR: Target[140.225.23.196_1107705856][_IN_] ' $target->[50]{$mode} ' did not eval into defined data
2015-03-04 15:07:31: ERROR: Target[140.225.23.196_1107705856][_OUT_] ' $target->[50]{$mode} ' did not eval into defined data
2015-03-04 15:07:31: ERROR: Target[140.225.23.195_1107705856][_IN_] ' $target->[101]{$mode} ' did not eval into defined data
2015-03-04 15:07:31: ERROR: Target[140.225.23.195_1107705856][_OUT_] ' $target->[101]{$mode} ' did not eval into defined data
2015-03-04 15:07:31: ERROR: Target[140.225.22.7_5179][_IN_] ' $target->[104]{$mode} ' did not eval into defined data
2015-03-04 15:07:31: ERROR: Target[140.225.22.7_5179][_OUT_] ' $target->[104]{$mode} ' did not eval into defined data
2015-03-04 15:07:31: ERROR: Target[140.225.22.6_5179][_IN_] ' $target->[116]{$mode} ' did not eval into defined data
2015-03-04 15:07:31: ERROR: Target[140.225.22.6_5179][_OUT_] ' $target->[116]{$mode} ' did not eval into defined data

If they are not valid anymore, try deleting them in XI and removing the rrd files too.
Keep running this until the errors are gone.

Code: Select all

sudo LANG=C LC_ALL=C /usr/bin/mrtg /etc/mrtg/mrtg.cfg

When it is error free, restart cron

Code: Select all

service crond restart