BIND 9 Memory Leak?

[Greg Burch]

Stephen, thanks for your response.  Our servers are serving as caching
name servers and also serving up internal-only zones.  We have
thousands of clients querying each name server.

You're correct in that we compiled the versions of BIND ourselves, but
we did not tweak any compiler options from the default.

I'm not sure I understand exactly what you're driving at with the "look
for data that doesn't match your server" suggestion...there definitely
will be cached Internet data there.  The issue I had was that the zone
files and cache dump added up to 7.7M at that moment in time, yet the
named process is using 391M of memory.  That's a very large
discrepancy.

Since we've begun doing a nightly restart of the named process, we have
not experienced the swapping issues, so it definitely looks like a
memory leak of some kind from where I'm sitting.

Thanks,
Greg

On Jan 23, 3:57 pm, "Stephen John Smoogen" <smo... at gmail.com> wrote:
> On 23 Jan 2007 06:36:56 -0800, Greg Burch <greg.bu... at gmail.com> wrote:
>
> > We're running multiple name BIND servers (on Red Hat EL AS4) in our
> > organization (some at version 9.3.1, others have been upgraded to
> > 9.3.2-P2), and in the last month or two, we've been running into issues
> > where DNS resolution on some of our servers slows to a crawl, because
> > the memory begins swapping to disk.  We've tried various things, such
> > as tweaking the max-cache-size and cleaning-interval parameters, with
> > no luck.  There had been no recent upgrades to BIND before these
> > problems began occurring, although other packages on the system may
> > have been updated via a Red Hat satellite server.Are the servers acting as both caching/resolving name servers for
> other machines and as an authoritative name server for your zone? How
> many clients are getting the data?
>
> The other issue might be with compile options and choices when you
> compiled your version of BIND. The RHEL-4 Bind is 9.2.4, so I am
> guessing there was a recompile somewhere.
>
> > As an example of the "memory leak", here's top output that shows BIND
> > using over 300M of memory:
>
> > PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
> > 18246 named     19   0  391m 334m 1204 S   17 33.1   1462:47 named
>
> > When I run an "rndc dumpdb", we can see the cache is 5.8M:
>
> > -rw-r--r--  1 named named 5812840 Jan  5 14:31 named_dump.dbThe file on my part is an ASCII dump so look for data that doesnt
> match your server.
>
> > The data itself that the server is master for is 1.9M.--
> Stephen J Smoogen. -- CSIRT/Linux System Administrator
> How far that little candle throws his beams! So shines a good deed
> in a naughty world. = Shakespeare. "The Merchant of Venice"