SUMMARY : unbearable DNS time outs recently.

From: chas <panda_at_peace.com.my>
Date: Sun, 03 May 1998 17:06:50 +0800 (SGT)

Sorry, forgot to add the "SUMMARY" in last post.
(doing so for the archive).

---
Sincere thanks to Dejan Muhamedagic and Richard Eisenman
for ideas concerning my DNS problems. I think that DNS
and the Alpha 500 were not at fault... but we suspect 
problems on the router. I set up DNS on a new 
machine with the latest BIND and cache files and it was
also having timeout problems.  The problem cured itself within
12 hours and coincidentally, international network speed
between malaysia and the US/world increased a great deal...
probably an upstream problem.
Still, some good tips learned for debugging this and
thanks to Dejan and Richard once again,
chas
Dejan Muhamedagic :
-------------------
>Quite interesting, but I don't have a clue.  Still, have you tried
>sending SIGUSR1 to named and checking /var/tmp/*named* logs to see
>what's going on.  Running tcpdump and filtering out everything but
>domain port may be helpful too.
>
>I would do the following:
>
>Get two terminals and run nslookup in one and tcpdump port domain in
>another simultaneously and try to figure out from that.
>
>Also, if you have some spare time you may try upgrading to newer named
>release (8.1 or so).
>
>Good luck!
>
>Dejan
>
>chas wrote:
>....
>> Any way to monitor just how many queries named is answering and
>> tracking the bottleneck ?
>
>sending SIGUSR1 one or more times.
From: "Richard Eisenman" 
------------------------
Is your cache file up to date?  Does your system log indicate any repetitive
errors or problems with named transfers that may be dragging down the
daemon?

Received on Sun May 03 1998 - 11:26:00 NZST