Antwort: [Nagiosplug-help] Check_dig returning warning state incorrectly

srunschke at srunschke at
Mon Sep 26 01:01:02 CEST 2005


> I tried going back to the version in plugins 1.3.1 but unless I am 
> mistaken that version does not allow you to set warning and critical
> values for how long the dns takes to respond. This is something I 
> need though since on rare occasions (cough) one of the Windows boxes
> running DNS gets a little ... ehm ... sluggish. A situation which 
> impacts performance all around for scores of applications. 
> Incidentally I am running all this on a Quad-CPU smp kernel 2.6.9 
> Red Hat EL 4 box. Scanning through recent messages that seems to be 
> important too.
> Any suggestions are welcome.

this is a RHES4 issue.
Obviously they screwed up something in the futex handling kernelwise.

I suspect you are using kernel 2.6.9-11, possible workarounds are:

1. downgrade to 2.6.9-5 and the problems will be gone
2. download the bind9.src and run ./configure omitting --enable-threads 
and use
    the dig command you just made

Those problems arise with every bind9 utility, so using nslookup will not 
help your
problem since both dig and nslookup use the isc-lib from bind.

I am currently working with RedHat support on this issue and engineering 
is certain
to be able to release a hotfix kernel soonish.


Sascha Runschke
Netzwerk Administration

Robert-Bosch-Str. 1
40668 Meerbusch

Tel.:+49 (0) 2150.9153.226
Mobil:+49 (0) 173.5419665
mailto:SRunschke at
Sicherheitshinweis zur E-Mail Kommunikation /
  Security note regarding email communication:

More information about the Help mailing list