nsd users - Apr 2012 - nsd init patch

Hello,

on the big installations NSD can stop some secs/mins and "nsdc
restart"
fail.

this patch for nsd.init fix this:

--- /etc/init.d/nsd     2012-04-09 18:01:39.446909907 +0400
+++ /etc/init.d/nsd     2012-04-09 18:01:24.248314446 +0400
@@ -71,6 +71,7 @@
      # ensure notifies are sent at startup
      rm -f $NSD_XFRDFILE
   fi
+  while [[ $(pidof nsd) ]]; do sleep 1;done
   echo
   return $RETVAL
 }


-- 
Andrey Scopenco
http://www.mchost.ru

On Mon, 9 Apr 2012, Andrey Scopenco wrote:
> on the big installations NSD can stop some secs/mins and "nsdc
restart"
> fail.
>
> this patch for nsd.init fix this:
>
> --- /etc/init.d/nsd     2012-04-09 18:01:39.446909907 +0400
> +++ /etc/init.d/nsd     2012-04-09 18:01:24.248314446 +0400
> @@ -71,6 +71,7 @@
>      # ensure notifies are sent at startup
>      rm -f $NSD_XFRDFILE
>   fi
> +  while [[ $(pidof nsd) ]]; do sleep 1;done
>   echo
>   return $RETVAL
> }
That's a nice possible infinite loop there.

What is taking so long for you? The "nsd patch" should be blocking,
so it is the actual "stop" that is taking so long? Is it a very
busy server?

That said, I do experience problems with nsd and rhel init
scripts too, where a "service nsd stop" won't kill nsd,
but I believe those are related to nsdc reloads and pid
changing, and nsd being unable to deal with the pid file.

Paul