BIND 8.2.1 Hangs - NDC fails

Barrett Richardson barrett at phoenix.aye.net
Mon Sep 13 19:10:04 UTC 1999



On 13 Sep 1999, Eric Molitor wrote:
>
> I have the same problem here at work. Any chance you could post that script to
> the newsgroupor make it available?
> 
> - Eric Molitor
> 

Doens't do any error checking on the reload -- named could be dead
and the script wouldn't know. In my case I have another means that
monitors if it is running.  At first glance it looks like the problem
starts with logging. First logging ceases, but named continues to
answer requests, for a few minutes. When it stops answering requests,
an "ndc reload" will wake it back up and get logging going again.
This box is a Linux Alpha box and is being replaced with a Sun box this
afternoon. If it does the same on the Sun I'll send it a SIGSEGV to
entice it to dump core (see what it is trying to do when it's hung).

------------ cut ------------
#!/bin/sh

while :
do
  nslookup -timeout=5 127.0.0.1 127.0.0.1 > /dev/null 2>&1
  if [ $? -ne 0 ];
  then
      ndc reload > /dev/null 2>&1
      echo "ndc reload executed" | mail some.body.that.cares
   fi
   sleep 15
done
---------- cut ----------------- 


-

Barrett



More information about the bind-users mailing list