Maintained by: NLnet Labs

[Unbound-users] unbound: fatal error: services/outside_network.c:

W.C.A. Wijngaards
Fri Sep 28 15:29:14 CEST 2012


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Leo,

Sounds like a bug fixed in svn trunk. (first report for that was from
a Windows user, by the way).

Best regards,
   Wouter

On 09/28/2012 03:14 PM, Leo Bush wrote:
> Dear all,
> 
> We use unbound since more than a year already with great
> satisfaction and no issue for resolving 4 Mbps DNS traffic in
> average with DNSSEC enabled.
> 
> We used to have unbound 1.4.14. At the end of August suddenly
> unbound stopped to work. The process killed itself with the
> following message in the logs:
> 
> /var/log/messages:Aug 28 19:35:22 hostname unbound: [7896:1] fatal 
> error: services/outside_network.c:1504: serviced_callbacks:
> pointer whitelist fptr_whitelist_serviced_query(p->cb) failed
> 
> The machine was under no unusual CPU/Memory/Network load.
> 
> We thought, bad luck, started the process and everything was fine
> for one week, when suddenly again the same thing. Fatal error.
> 
> /var/log/messages:Sep  6 14:20:20 hostname unbound: [1503:0] fatal 
> error: services/outside_network.c:1504: serviced_callbacks:
> pointer whitelist fptr_whitelist_serviced_query(p->cb) failed
> 
> Then we thought about our version being too old and we upgraded to
> the latest release unbound 1.4.18.
> 
> Now we are running unbound 1.4.18 since two weeks and today again:
> 
> /var/log/messages:Sep 28 13:59:53 hostname unbound: [1591:1] fatal 
> error: services/outside_network.c:1510: serviced_callbacks:
> pointer whitelist fptr_whitelist_serviced_query(p->cb) failed
> 
> 
> Sounds like a software bug?
> 
> kind regards
> 
> Leo Bush
> 
> 
> By the way, even if I do not think it is related to this case, I
> would like to report on the following behavior, that we note: 
> Several times a day we notice short peaks (<10 minutes) in
> outgoing traffic (times 7). Even if the traffic is so much higher,
> we do not notice a performance issue on the server. CPU and Ram are
> still fine. DNS response times are fine too. A traffic analysis
> suggested a DNS amplification attack where the question is for ANY
> isc.org. This attack never was happening when unbound crashed.
> 
> 
> _______________________________________________ Unbound-users
> mailing list Unbound-users at unbound.net 
> http://unbound.nlnetlabs.nl/mailman/listinfo/unbound-users

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://www.enigmail.net/
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=pB4i
-----END PGP SIGNATURE-----