bind server dying

[Joseph S D Yao]

> I was awaken by my boss this morning to discover that all three of our DNS
> server processes were down.  Upon looking at the DNS log file, they all died
> with the following error message:
> 
> 14-Oct-1999 06:51:11.382 insist: critical: memcluster.c:307: INSIST(stats[size].gets != 0) failed.
> 14-Oct-1999 06:51:11.382 panic: critical: memcluster.c:307: INSIST(stats[size].gets != 0) failed.
> 
> Does anyone know what these messages mean, and what might have caused the
> problem?  I find it suspicious because I've been running these servers for
> several months now and they've never had any problems.  Than all of a sudden,
> all three of them die within a two hour time span.  Could it possibly be a
> security hole that killed the named process?
> 
> By the way, we are using bind 8.2
> 
> Thank you for any help, Steven

Didn't you just ask this a couple of hours ago?

The "patch1" [there is no "patch2"] on the ISC server fixes this.  So
does upgrading to 8.2.1 - unfortunately, the patches for that are only
available by perusing the archives of this mailing list.

--
Joe Yao				jsdy at cospo.osis.gov - Joseph S. D. Yao
COSPO/OSIS Computer Support					EMT-B
-----------------------------------------------------------------------
This message is not an official statement of COSPO policies.