[Unbound-users] fatal error unbound (fatal error: services/outside_network.c:1504: serviced_callbacks: pointer whitelist fptr_whitelist_serviced_query(p->cb) failed)

[W.C.A. Wijngaards]

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

Hoi Michiel,

On 06/27/2012 09:49 PM, M. Piscaer wrote:
> Hi,
> 
> At 17:39 today we restarted the service unbound with
> "unbound-control reload". At 18:54 we got an fatal error:
> 
> Jun 27 18:54:00 isp-nscache-01 unbound: [1733:0] error: validator:
> bad event module_event_reply Jun 27 18:54:00 isp-nscache-01
> unbound: [1733:0] fatal error: services/outside_network.c:1504:
> serviced_callbacks: pointer whitelist
> fptr_whitelist_serviced_query(p->cb) failed
> 
> after that we needed to start the service:
> 
> Jun 27 19:12:42 isp-nscache-01 unbound-anchor:
> /var/lib/unbound/root.key has content Jun 27 19:12:42
> isp-nscache-01 unbound-anchor: success: the anchor is ok Jun 27
> 19:12:42 isp-nscache-01 unbound: [12534:0] warning: did not exit
> gracefully last time (1733) Jun 27 19:12:42 isp-nscache-01 unbound:
> [12535:0] notice: init module 0: validator Jun 27 19:12:42
> isp-nscache-01 unbound: [12535:0] notice: init module 1: iterator 
> Jun 27 19:12:42 isp-nscache-01 unbound: [12535:0] info: start of
> service (unbound 1.4.16).
> 
> What happend and what can we do that is not happen again?

There are two failures here.  From reading back the code I have not
found the bug.  The bad event failure is that the validator complains
that answers from the network happen, which is impossible, and flags
this as an internal error.  The second failure is that the contents of
callback->next seems to contain garbage, and an assertion fails
because of the garbage.  The second failure happens at the same time
as the first; because otherwise an assertion on line 1519 should fire.
 Did you compile with --enable-debug?  If not, could you do so?  The
additional assertions may provide more information should the bug
reappear, and it is not heavy like enableing high verbosity.

> The unbound service is handeling 2044 query's per seconde. We are
> running unbound version 1.4.16?

Nice throughput.  Do you have any special modules configured (special
python scripts?) ?

> Google-ing on this issue doesn't give any result.

This is the first report of this bug.  I would like to be able to
reproduce this problem.  Do you have query-logging perhaps and can
tell me what queries were active in the last second(s) ?

You seem Dutch, and so is NLnet Labs, in case language makes bug
reports better :-)

Best regards,
   Wouter
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=HNuX
-----END PGP SIGNATURE-----