[debian-lan-devel] DNS causing issues in network performance

[Andreas Mundt]

On October 11, 2016 3:34:20 AM GMT+03:00, BMIRC System Administrator <bmi.sysadmin at mail.sdsu.edu> wrote:

>I am using DebianLAN on the cluster for my lab, and the server went
>down a
>few days ago. Once I rebooted the head node, all compute nodes and
>workstations came back up, but ever since then all workstations have
>been
>very slow, have had intermittent issues connecting to the internet
>(usually
>with DNS errors), and login will seemingly randomly fail about
>one-third of
>the time.
>
>I suspect the issue is with DNS. Do you have a hypothesis as to what
>the
>problem might be?

No idea.   :-/  

>I watched the syslog as client workstations tried to log in, and the
>folllowing error message appeared:
>
>icinga: SERVICE ALERT: workstation00;Zombie
>Procs;CRITICAL;HARD;4;CHECK_NRPE: Socket timeout after 10 seconds

This is just a message from icinga, probably not related (or just a symptom) of the problem.

Can you test dns? Perhaps forwarders need to be added to bind? 

Best regards,

Andi


Hi Sam,
-- 
Sent from mobile, excuse my brevity.