[Unbound-users] SIGSEGV w/ unbound 1.1.1 (compiled from ports) and FreeBSD 7.0-RELEASE-p7
Sven Juergensen (KielNET)
s.juergensen at kielnet.de
Mon Jan 5 06:48:51 UTC 2009
-----BEGIN PGP SIGNED MESSAGE-----
W.C.A. Wijngaards wrote:
> Hi Sven,
> You are using libevent, I think it is too old. I know version 1.1 has
> trouble (everywhere?), 1.3 seems to work fine on linux(epoll).
> Anyway, I think libevent-1.4.8-stable has no problems whatsoever (no
> problems reported to me up to now).
> What version of libevent are you using? Can you upgrade it and tell me
> if things change for you?
pkg_info reports libevent-1.4.8
> You can also compile without libevent which also stops the problem.
I will try that if things go south again.
Not sure if this is related to the problem
but after I restored the config as it was
when trying the server from home, things
apparently work again but I shall monitor
this for some more. I disabled the extended-
statistics and threading.
BTB, on the unbound.net site it says you can
tune unbound's memory usage. The machine I
have unbound running on is a quadcore xeon
with 8GByte of RAM. According to the comments
in the unbound.conf, is the following reasonable
and can unbound use that much?
> I think there may need to be a build dependency on libevent version >=
> 1.4.8 on unbound, since you are the 3rd or 4th guy to report this. This
> could get tricky, since I try to remain compatible with
> libevent-compatible replacement libraries, such as libev. (
> http://software.schmorp.de/pkg/libev.html )
> Best regards,
> Wouter (and a Happy New Year)
> Sven Juergensen (KielNET) wrote:
>> Hi list,
>> happy new year and stuff ;)
>> I'm in the progress of replacing some
>> old tinydns servers with unbound/NSD
>> primarily for built-in IPv6 support
>> and performance reasons.
>> It worked okay when I was testing it
>> from home with a single user. This
>> morning I put a little more load onto
>> it by integrating into our intranet
>> to do some tests before deploying it
>> for our customers.
>> This is where it started to signal 11
>> when queries began. Currently, there
>> are only IPv4 queries.
>> pid 569 (unbound), uid 59: exited on signal 11
>> This is reproducible so I went and did
>> a ktrace if that helps. I'm no coder
>> and I can't really make sense of that
>> output. If you need me to put this some-
>> where, just let me know.
>> BTW: is it normal, that the unbound pro-
>> cess is in the kqread state?
>> Any input on how to track this down fur-
>> ther would be appreciated. If any other
>> information is needed, just let me know.
>> Many thanks and best regards,
>> Sven Juergensen
>> The config looks like this:
>> verbosity: 1
>> statistics-interval: 3600
>> extended-statistics: yes
>> interface: 0.0.0.0
>> interface: ::0
>> access-control: 0.0.0.0/0 allow
>> access-control: ::0/0 allow
>> chroot: "/usr/local"
>> username: unbound
>> directory: "/usr/local/etc/unbound"
>> logfile: "/usr/local/var/log/unbound.log"
>> root-hints: "named.cache"
>> hide-identity: yes
>> hide-version: yes
>> Mit freundlichen Gruessen,
>> i. A. Sven Juergensen
> Unbound-users mailing list
> Unbound-users at unbound.net
Unbound-users mailing list
Unbound-users at unbound.net
Mit freundlichen Gruessen,
i. A. Sven Juergensen
Gesellschaft fuer Kommunikation
Preusserstr. 1-9, 24105 Kiel
Telefon : 0431 / 2219-053
Telefax : 0431 / 2219-005
E-Mail : s.juergensen at kielnet.de
Geschaeftsfuehrer Eberhard Schmidt
HRB 4499 (Amtsgericht Kiel)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)
-----END PGP SIGNATURE-----
More information about the Unbound-users