Tuesday, September 19, 2017

Re: Crash on stopping relayd

Hi Maxim,

yeah I'll wait till 6.2 and do a upgrade. I misstyped the OS version I
have a OpenBSD 6.1 running. I will be just a little more careful for now
though :)

Regards

Markus

Am 19.09.2017 um 22:09 schrieb Maxim Bourmistrov:
>
> Checked out http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys/net/pf.c <http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys/net/pf.c>
> Looks like this bug is fixed after 6.1 , e.g. in 6.2.
>
>> 19 sep. 2017 kl. 18:43 skrev rosjat <rosjat@ghweb.de>:
>>
>> Hi there,
>>
>> I don't know if someone had also a problem like this. What happen is:
>>
>> I was stopping the relayd daemon with rcctl stop relayd and lost connection to my remote system. An hour later, sitting in front of the server in the data center I saw that openbsd crashed. That what I could see in the logs.
>>
>> /bsd: panic: kernel diagnostic assertion "(sk->inp == NULL) || (sk->inp->inp_pf_sk == NULL)" failed: file "/usr/src/sys/net/pf.c", line 7072
>> Sep 19 13:59:42 heimdal /bsd: Stopped at Debugger+0x9: leave
>> Sep 19 13:59:42 heimdal /bsd: TID PID UID PRFLAGS PFLAGS CPU COMMAND
>> Sep 19 13:59:42 heimdal /bsd: Debugger() at Debugger+0x9
>> Sep 19 13:59:42 heimdal /bsd: panic() at panic+0xfe
>> Sep 19 13:59:42 heimdal /bsd: __assert() at __assert+0x25
>> Sep 19 13:59:42 heimdal /bsd: pf_state_key_unref() at pf_state_key_unref+0xc6
>> Sep 19 13:59:42 heimdal /bsd: pf_pkt_unlink_state_key() at pf_pkt_unlink_state_key+0x15
>> Sep 19 13:59:42 heimdal /bsd: m_free() at m_free+0xa0
>> Sep 19 13:59:42 heimdal /bsd: sbdrop() at sbdrop+0x80
>> Sep 19 13:59:42 heimdal /bsd: sbflush() at sbflush+0x1f
>> Sep 19 13:59:42 heimdal /bsd: sbrelease() at sbrelease+0x11
>> Sep 19 13:59:42 heimdal /bsd: sorflush() at sorflush+0x158
>> Sep 19 13:59:42 heimdal /bsd: sofree() at sofree+0xa7
>> Sep 19 13:59:42 heimdal /bsd: soclose() at soclose+0xf3
>> Sep 19 13:59:42 heimdal /bsd: soo_close() at soo_close+0x1c
>> Sep 19 13:59:42 heimdal /bsd: fdrop() at fdrop+0x2c
>> Sep 19 13:59:42 heimdal /bsd: end trace frame: 0xffff80002526dda0, count: 0
>> Sep 19 13:59:42 heimdal /bsd: https://www.openbsd.org/ddb.html describes the minimum info required in bug
>> Sep 19 13:59:42 heimdal /bsd: reports. Insufficient info makes it difficult to find and fix bugs.
>>
>>
>> I run a OpenBSD 6.0, at that time I didn't had all syspatches installed ( till patch 015_sigio) so I don't know it it fixed after installing all patches.
>>
>> So it seems the problem isn't in the relayd daemon but in the pf, and maybe the relayd anchors?
>>
>> I feel a little reluctant right now to do anything with relayd on that machine because it will cause me just pain if the system crashes again. so if someone knows if this issue is fixed with an full patched system it would help a lot. And as the name suggest its a firewall machine and my SPOF :-(
>>
>> Regards
>>
>> --
>> Markus Rosjat fon: +49 351 8107223 mail: rosjat@ghweb.de
>>
>> G+H Webservice GbR Gorzolla, Herrmann
>> Königsbrücker Str. 70, 01099 Dresden
>>
>> http://www.ghweb.de
>> fon: +49 351 8107220 fax: +49 351 8107227
>>
>> Bitte prüfen Sie, ob diese Mail wirklich ausgedruckt werden muss! Before you print it, think about your responsibility and commitment to the ENVIRONMENT
>>
>
>

--
Markus Rosjat fon: +49 351 8107223 mail: rosjat@ghweb.de

G+H Webservice GbR Gorzolla, Herrmann
Königsbrücker Str. 70, 01099 Dresden

http://www.ghweb.de
fon: +49 351 8107220 fax: +49 351 8107227

Bitte prüfen Sie, ob diese Mail wirklich ausgedruckt werden muss! Before
you print it, think about your responsibility and commitment to the
ENVIRONMENT

No comments:

Post a Comment