New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
crash on run_rr_callbacks() from dlg_onroute() #1059
Comments
Any idea why frames 1 and 2 don't have the symbols? Was there a single core file, presuming the options for core files per pid was set? |
No idea, per-process core file is enabled too.
|
same thing in another customer
|
Can you do:
The code in branch 4.4 for dlg_handlers.c:1350 is:
So no strlen() to be used there. |
This is the In order you to see the source easily I pasted dlg_handlers.c already patched at https://paste.debian.net/926240/
So relevant code is too
but the I would say the problem is |
Did you put in pastebin the wrong file (like dlg_db_handlers.c)? All the tokens provided by kamailio in that log message are by length and pointer (%.*s), so there should be no strlen done on the pointers we set, so not sure on what syslog can do the strlen and crash. |
I wonder if there is a copy/paste issue or the
After event and at the end. First looks like a valid value, second a bit more strange. Can you get in the same frame:
|
Yes, I did, sorry. The right file at https://paste.debian.net/926247/ |
|
for the relevant part of code
|
and we have this at the log at the time of the core, 11064 is another process.
|
Is it possible that there is a time consuming operations for BYE (eg, a slow database operation -- accounting or something else) that can take more than 5 seconds? |
more than 5 secs? I don't think so. @apogrebennyk can you confirm? |
We have latency logging enabled but there were no abnormally long delays reported. The last lines before the crash are like this:
|
- done for dlg_onroute(), as an extra safety for early detection of races, related to GH #1059
kamailio/kamailio#1059 Change-Id: I3dcad249537f92c2ef75708a60f53b5b5383264a
kamailio/kamailio#1059 Change-Id: I3dcad249537f92c2ef75708a60f53b5b5383264a (cherry picked from commit 1e2115a)
kamailio/kamailio#1059 Change-Id: I3dcad249537f92c2ef75708a60f53b5b5383264a (cherry picked from commit 1e2115a)
kamailio/kamailio#1059 Change-Id: Ic0b3e5b84d92c0a2348a839fa7915597ea501605
kamailio/kamailio#1059 Change-Id: Ic0b3e5b84d92c0a2348a839fa7915597ea501605 (cherry picked from commit f81cba7)
kamailio/kamailio#1059 Change-Id: I3dcad249537f92c2ef75708a60f53b5b5383264a (cherry picked from commit 1e2115a) (cherry picked from commit 026743a)
kamailio/kamailio#1059 Change-Id: I3dcad249537f92c2ef75708a60f53b5b5383264a (cherry picked from commit 1e2115a)
If you haven't ported the patches referenced above yourself, I just did it. Can you try with latest 4.4 branch and see the results? |
…ute execution - avoid execution of script callbacks for event route - related to GH kamailio#1059 and kamailio#1069
Closing this one, eventually fixed with the patches from the few weeks ago. If issue persists, re-open. |
It seems that kamailio received two BYE messages at the same time and dlg was gone?
Debugging Data
Log Messages
Additional Information
kamailio -v
This is
NGCP
based on 4.4.4 with tm deep cloning edb9df8 backportedThe text was updated successfully, but these errors were encountered: