[mrtg] data collection did not complete within interval

[Clark Kee]

Yes, I've reviewed the online documentation on this subject (and the 
applicable chapter of Steve's book). We've increased the Forks parameter 
from 2 to 4 to 18, to no avail. Can someone comment on the different 
performance implications of increasing the Forks param vs. running 
multiple mrtg instances. This is running on an ultrasparc with 23 
virtual processors, and load average around 2.

There are errors such as the following:
2010-08-18 15:23:54 -- 2010-08-18 14:54:54: ERROR: 
Target[sometarget][_IN_] ' $target->[671]{$mode}  * 1.8 + 32' (warn): 
Use of uninitialized value in multiplication (*) at (eval 1344) line 1.

Am I reading this correctly that this failing query took 29 minutes? 
What would prevent that from timing out? The mrtg snmp libraries use the 
default timeout/retry/backoff parameters.

C.

Clark Kee wrote:
> Our mrtg/rrd installation, which has been running fine for months, 
> recently stopped being able to produce most or all graphs. Inpection 
> of the rrd files showed NaNs were being written, and the mrtg log 
> showed the warning in the subject line of this message. We were 
> running with 2 forks. Increasing to 4 did not seem to help. We monitor 
> a little over 800 targets. Any suggestions as to how to deal with 
> this, or additional diag info to gather welcome.
>
> mrtg 2.16.2
> rrdtool 1.3.1
>
> Clark Kee
> PINS
>