[OpenSIPS-Devel] OpenSIPS Crash

Ben Newlin Ben.Newlin at genesys.com
Fri Jun 7 09:27:13 EDT 2019


Bogdan,

I no longer have the original backtrace I posted in May, but if it was the same issue then it has only happened those two times, both when under load. I have not been able to reproduce it reliably or with single calls.

For the double ACK, so if the MF value is changed then it is not just traced twice it was actually sent twice. But what scenario would cause retransmission of a hop-by-hop ACK? I can’t think of one. So it still seems strange. But it may be a symptom of the issue and not a cause.

Ben Newlin

From: Bogdan-Andrei Iancu <bogdan at opensips.org>
Date: Friday, June 7, 2019 at 9:15 AM
To: Ben Newlin <Ben.Newlin at genesys.com>, OpenSIPS devel mailling list <devel at lists.opensips.org>
Subject: Re: [OpenSIPS-Devel] OpenSIPS Crash

Hi Ben,

How often/easy is to reproduce this crash (if possible) ? Brainstorming with Razvan, we suspect a race (on the msg save in shmem in transaction) between the process doing the cleanup after the async resume and the process running the failure route (due th 503).

But this is just a supposition, eventually you can validate it or not by removing the async ??

And on the double ACK - I'm not 100% it is a actually a double one, as the second has a smaller MF value (69, versus the 70 on the first ACK).

Regards,


Bogdan-Andrei Iancu



OpenSIPS Founder and Developer

  https://www.opensips-solutions.com<https://www.opensips-solutions.com>

OpenSIPS Summit 2019

  https://www.opensips.org/events/Summit-2019Amsterdam/<https://www.opensips.org/events/Summit-2019Amsterdam/>
On 06/07/2019 03:52 PM, Ben Newlin wrote:
Bogdan,

Sorry, I should have thought to actually look at the trace and examine this call.

1) Yes
2) The Called Party is 10.32.20.60, which is another OpenSIPS instance. The crashed instance received the "503 Service Unavailable" approximately 8-10 ms after sending the INVITE.

There is a SIP trace of the exchange here: https://pastebin.com/6bttsSVD<https://pastebin.com/6bttsSVD>.

One oddity I saw is that the crashed process appears to send (or at least siptrace) the ACK twice.

Ben Newlin

From: Bogdan-Andrei Iancu <bogdan at opensips.org><mailto:bogdan at opensips.org>
Date: Thursday, June 6, 2019 at 11:42 AM
To: OpenSIPS devel mailling list <devel at lists.opensips.org><mailto:devel at lists.opensips.org>, Ben Newlin <Ben.Newlin at genesys.com><mailto:Ben.Newlin at genesys.com>
Subject: Re: [OpenSIPS-Devel] OpenSIPS Crash

Hi Ben,

Thanks for "another" report :).

Questions:
1) do you do any async for the INVITE in this crash ?
2) if it is an YES to (1), is the caller party generating the "503 Service Unavailable" (which triggers the crash) - 10.32.20.60 ?? - a really close (from net delay perspective) and fast to answer party ?

Regards,



Bogdan-Andrei Iancu



OpenSIPS Founder and Developer

  https://www.opensips-solutions.com<https://www.opensips-solutions.com>

OpenSIPS Summit 2019

  https://www.opensips.org/events/Summit-2019Amsterdam/<https://www.opensips.org/events/Summit-2019Amsterdam/>
On 06/05/2019 10:02 PM, Ben Newlin wrote:
We have had another crash today.

Backtrace is here: https://pastebin.com/q4RQC7kS<https://pastebin.com/q4RQC7kS>

I found this in the log at the time of the crash:

Jun  5 17:54:10 [4978] CRITICAL:core:sig_usr: segfault in process pid: 4978, id: 8


Please let me know if any further information can be useful.

Ben Newlin

From: Devel <devel-bounces at lists.opensips.org><mailto:devel-bounces at lists.opensips.org> on behalf of Ben Newlin <Ben.Newlin at genesys.com><mailto:Ben.Newlin at genesys.com>
Reply-To: OpenSIPS devel mailling list <devel at lists.opensips.org><mailto:devel at lists.opensips.org>
Date: Friday, May 10, 2019 at 6:31 PM
To: OpenSIPS devel mailling list <devel at lists.opensips.org><mailto:devel at lists.opensips.org>
Subject: Re: [OpenSIPS-Devel] OpenSIPS Crash

I found this in the log at the time of the crash:

kernel: opensips[5003]: segfault at 30 ip 00007fbd4c8f59d0 sp 00007ffcaa850c80 error 6 in tm.so[7fbd4c887000+8e000]

Ben Newlin

From: Devel <devel-bounces at lists.opensips.org><mailto:devel-bounces at lists.opensips.org> on behalf of Ben Newlin <Ben.Newlin at genesys.com><mailto:Ben.Newlin at genesys.com>
Reply-To: OpenSIPS devel mailling list <devel at lists.opensips.org><mailto:devel at lists.opensips.org>
Date: Friday, May 10, 2019 at 5:44 PM
To: OpenSIPS devel mailling list <devel at lists.opensips.org><mailto:devel at lists.opensips.org>
Subject: [OpenSIPS-Devel] OpenSIPS Crash

Hello,

We had a crash today of our OpenSIPS instance.

Backtrace is here: https://pastebin.com/QbRJimwx<https://pastebin.com/QbRJimwx>

# opensips -V
version: opensips 2.4.5 (x86_64/linux)
flags: STATS: On, DISABLE_NAGLE, USE_MCAST, SHM_MMAP, PKG_MALLOC, F_MALLOC, FAST_LOCK-ADAPTIVE_WAIT
ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16, MAX_URI_SIZE 1024, BUF_SIZE 65535
poll method support: poll, epoll, sigio_rt, select.
git revision: d025b4f61
main.c compiled on 20:58:31 May  9 2019 with gcc 7

Ben Newlin





_______________________________________________

Devel mailing list

Devel at lists.opensips.org<mailto:Devel at lists.opensips.org>

http://lists.opensips.org/cgi-bin/mailman/listinfo/devel<http://lists.opensips.org/cgi-bin/mailman/listinfo/devel>





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/devel/attachments/20190607/137cc30b/attachment-0001.html>


More information about the Devel mailing list