still have named memory leak

Frank Bulk frnkblk at iname.com
Sat Dec 13 18:19:56 UTC 2014


Here’s some suggestions from ISC on capturing information on this memory growth issue:

https://kb.isc.org/article/AA-01208

 

Frank

 

From: bind-users-bounces at lists.isc.org [mailto:bind-users-bounces at lists.isc.org] On Behalf Of Kevin Oberman
Sent: Saturday, December 13, 2014 12:07 PM
To: Mukund Sivaraman
Cc: bind-users at lists.isc.org
Subject: Re: still have named memory leak

 

On Fri, Dec 12, 2014 at 8:12 AM, Mukund Sivaraman <muks at isc.org <mailto:muks at isc.org> > wrote:

Hi Len

On Fri, Dec 12, 2014 at 09:52:23AM -0600, lconrad at go2france.com <mailto:lconrad at go2france.com>  wrote:
> binary upgraded Freebsd 10 to Freebsd 10.1
>
> named  9.10.1, compiled from source
>
> at named start, 305 MB memory
>
> after several hours of running named is approaching 800 MB. I'm sure after a
> couple of days, as before, it will head towards 2000 MB
>
> suggestions?
>
> this is a recursive only NS, about 20M q/day restricted by ACL to
> "ournetworks"

This tells us that the named process size grows large, but more
information is needed to discover why. Can you send us the following?

1. Your named configuration.

2. Regular dumps over time of the statistics that are available via
HTTP, as the named process grows. See the "statistics-channels"
documentation in the manual. You can use curl or wget to dump them to a
file.

 

The standard tool for this on FreeBSD is fetch(1). E.g. fetch -o FILE "URL"

In a script I usually also use '-q' to eliminate noise, but YMMV.

 

I suspect most systems have wget and/or curl installed, but fetch is always present on FreeBSD.

--
R. Kevin Oberman, Network Engineer, Retired
E-mail: rkoberman at gmail.com <mailto:rkoberman at gmail.com>  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.isc.org/pipermail/bind-users/attachments/20141213/1ae2abf8/attachment.html>


More information about the bind-users mailing list