Bind 8.2.2 P5 hanging up...

[Philip L. Butler]

Kevin,

I am using Slackware, kernel version 2.0.38.  I will turn on 
debugging and see what it turns up.  I agree with you though in that 
I was looking at logs this afternoon from another system that is not 
hanging up and it showed the same cleaned cache message.  Do you know 
of a debug level to suggest ??  I have turned it on level 9, but I 
haven't had a chance to look at the logs since I did that.

BTW, on a system that has no problems, the LAN addresses are in the 
10.x.x.x range but the system that doesn't work is on a 101.0.0.x 
range - not a private IP range.  It was set up before I got involved. 
I wonder if that might have anything to do with it.  It shouldn't, 
but you never know.

Thanks,

Phil Butler

>Just because "Cleaned cache" is the last messages you see in the logs
>before the failure doesn't necessarily mean that it's related to the
>lock-up problem. Turn on debugging and that will give you a better idea
>of what it was doing prior to the failure.
>
>It does seem from the number of reports that there is some sort of
>chronic problem with 8.2.2-p5 on Linux (what flavor/version of Linux, by
>the way, you didn't specify?). If ISC isn't already actively on the
>trail of this, then maybe some debugging output might help track down
>the problem...
>
>
>- Kevin
>
>Philip L. Butler wrote:
>
>  > Hi,
>  >
>  > I am new to this list, so if this has been covered already I
>  > apologize.  I have scanned the list archives for help in this area -
>  > very useful information, but I am still puzzled a bit on a problem I
>  > am having.
>  >
>  > I am running a Linux box with Bind 8.2.2-P5.  Occasionally, it stops
>  > doing DNS for other machines.  From the logs, it seems (if I am
>  > interpreting them correctly) that it happens after the "Cleaned
>  > cache" message (see below) is logged and then subseqent DNS requests
>  > from the LAN result in the "Connection refused" message (see below).
>  > I have tried stripping all of my options out (listen-on and
>  > forwarders mainly) but that doesn't seem to help.  Rebooting helps
>  > for a while, but this obviously is not a permanent solution.
>  >
>  > I have scanned the OReilly DNS/Bind book (3rd edition) and there's an
>  > option to turn off clearing the cache, but it seems like I would want
>  > the Bind RAM cache to be pruned every so often.
>  >
>  > I haven't dug into the bind sources yet, but will be doing so soon.
>  >
>  > Has anyone run across this situation ??  In searching the bind-users
>  > archives for "ns_req", I see someone has had similar problems, but I
>  > haven't found the answer for my situation yet.
>  >
>  > Thanks in advance.
>  >
>  > Phil Butler
>  > butler at dii.net
>  >
>  > Logs
>  > ****
>  > Feb  3 15:48:08 (none) named[43]: Cleaned cache of 41 RRsets
>  > ...
>  > Feb  3 16:30:18 (none) named[43]: ns_req: sendto([172.23.9.2].137):
>  > Connection refused