I got a crash today in kamailio 4.1 (commit
1cc0144e434fdcf76013e9424d31928c260d4377) that was running fine in production
for a very long time.
I am working on upgrading to 4.4 but I am reporting this as a bug might exist
in latest code.
```
Core was generated by `/usr/local/src/git/kamailio-
I have tcpdump saving all SIP traffic to files. Based on the
Call-ID: 3a5e35a4-41f6-1235-7db0-000c29e59fdd
we can see at frame 5, I identified the INVITE that crashed kamailio as this
one:
​INVITE sip:[email protected]:5060 SIP/2.0
Via: SIP/2.0/UDP 202.173.5.217;rport;branch=z9hG4b
The kamailio logs at the time of the crash:
2016-12-21T16:59:06.281803+09:00 IPX182
/usr/local/src/git/kamailio-4.1/kamailio[11158]: :
[msg_translator.c:2316]: build_res_buf_from_sip_req():
BUG:build_res_from_sip_req: diff len=420 p-buf=415
2016-12-21T16:59:07.269816+09:00 IPX182
/usr/local/s
Closed #898.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/issues/898#event-905275684___
sr-dev mailing list
[email protected]
http:
I am closing this issue as I am planning to move to kamailio 4.4 so even if in
the meantime this happens again, there will be no follow up as I will consider
it a known issue in 4.1. Thanks.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or vie
### Description
JSON string returned by "kamctl ps" is malformed.
It seems it should end with:
}
but it ends with:
],
Reproduction
```
root@lab002139:~# kamctl ps
{"jsonrpc":"2.0","result":[{"IDX":0,"PID":19241,"DSC":"main process -
attendant"},{"IDX":1,"PID":19243,"DSC":"udp receive
I tested with a subsequent commit 55c941c052f0a8af8ba9c82ef75523dcd97e21a0 and
it is OK.
And yes, I was using default value 0 for pretty_format.
Thanks.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/kama
Closed #935.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/issues/935#event-928303947___
sr-dev mailing list
[email protected]
http:
OK. I have restarted tests (6 kamailio VMs on kvm, vmware and DigitalOcean)
with master branch latest commit 171553ebcac2f00a015191f36f2f228c0d3e483b.
---
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/issues/301#issuecomment-140050255___
One of the VM hosts in our lab got a hardware failure so i could not keep the
tests running.
I could not reproduce the problem with kamailio stopping processing messages.
But i got the same crash again in one of the remaining VMs in the lab:
Core was generated by `/usr/local/src/git/kamailio-4.4/
The problem of having kamailio stopping processing messages happened again:
```
[root@lab002185-flip-server kamailio-4.4]# kamctl fifo version
Server:: kamailio (4.4.0-dev4 (x86_64/linux))
Build:: mi_core.c compiled on 20:23:03 Sep 14 2015 with gcc 4.4.7
Flags:: STATS: Off, USE_TCP, USE_TLS, USE_SC
The problem of having kamailio stopping to process messages happened in one of
the DigitalOcean VMs. I can give you access to it if you want. Just pass me the
email to where i should send the login info.
---
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/
This is an intermittent problem (it doesn't happen all the time).
Version:
```
[root@lab002189-flip-server install]# kamctl fifo version
Server:: kamailio (4.3.3 (x86_64/linux))
Build:: mi_core.c compiled on 17:06:31 Oct 5 2015 with gcc 4.4.7
Flags:: STATS: Off, USE_TCP, USE_TLS, USE_SCTP, TLS_HO
Commit is the latest for 4.3: 48d78ce87e2f42bdc77e688a0f45ff725ea97de5
Here is the backtrace:
```
Program terminated with signal 11, Segmentation fault.
#0 0x7f3b4fd779da in update_dialog_vars_dbinfo (cell=0x7f3b3c3aa958,
var=0x3433) at dlg_db_handler.c:651
651 SET_STR_VALUE(valu
Here it is:
```
(gdb) frame 0
#0 0x7f3b4fd779da in update_dialog_vars_dbinfo (cell=0x7f3b3c3aa958,
var=0x3433) at dlg_db_handler.c:651
651 SET_STR_VALUE(values+2, var->key);
(gdb) p *cell
$1 = {ref = 1, next = 0x0, prev = 0x0, h_id = 10067, h_entry = 3273, state = 1,
lifetime = 1
```
(gdb) p *cell->vars
$5 = {key = {s = 0x7f3b3c43e7c8 "answer_time", len = 11}, value = {
s = 0x7f3b3c553ea0 "1440488615.120", len = 14}, vflags = 0,
next = 0x7f3b3c35c678}
(gdb) set $p=cell->vars->next
(gdb) while ($p != 0)
>p *$p
>set $p=$p->next
>end
$6 = {key = {s = 0x7f3b3c4064f0 "
There are errors in the log, but not for the callid mentioned in the bt.
Here is the full content of kamailio.log:
```
[root@lab002185-flip-server ~]# cat /var/log/kamailio/kamailio.log
2015-08-25T09:27:02.220867+09:00 lab002185-flip-server
/usr/local/src/git/kamailio-4.3/kamailio[911]: WARNING:
Yes. Only one core file. I was running a load test with sipp. When kamailio
stopped replying, i found the core file. So I don't think it got overwritten
unless it is possible for two processes to segfault at the same time and dump
to the same file name.
```
(gdb) p process_no
$1 = 1
(gdb) p pt[
OK, I am testing master too.
I have started the load tests. I should have results within 24 hours. However,
latest commit 8d8e1b957d9461d80b15b8e52464322a4e2b5fb1 of master is logging
this:
2015-08-27T14:36:07.034018+09:00 lab002187-flip-server
/usr/local/src/git/kamailio-devel/kamailio[1958]:
Here's what I got using 3 VMs:
1) CentOS6.6, 1CPU, 1GB, kamailio 4.3 commit
48d78ce87e2f42bdc77e688a0f45ff725ea97de5:
Crashed in update_dialog_vars_dbinfo
2) CentOS6.6, 1CPU, 2GB, kamailio 4.3 commit
48d78ce87e2f42bdc77e688a0f45ff725ea97de5
Crashed in update_dialog_vars_dbinfo generating 2 cor
I got a new crash in another small VM.
There was some corruption as the linked list got into a loop:
```
Core was generated by `/usr/local/src/git/kamailio-4.3/kamailio -m 256 -M 16 -P
/var/run/kamailio.pid'.
Program terminated with signal 11, Segmentation fault.
#0 0x7f8f2a2aa9da in update_d
> Interesting that it is same invalid value var=0x3433, but dlg->vars list is
> ok.
> Is v4.1 saving the dlg vars in db?
Yes. It is exactly the same cfg file (I had only to adjust DB tables when
moving to to 4.3).
> Can you describe a bit when the crash is happening? Like: if you run a stress
> Can you give the local variables in frame 1:
I assume you are talking about the last crash i mentioned with corruption of
linked list:
```
(gdb) frame 1
#1 0x7f8f2a2ab511 in update_dialog_dbinfo_unsafe (cell=0x7f8f1634b738) at
dlg_db_handler.c:718
718 if (upda
Here is the gdb output for crash using commit
2962852bf706692bdbc9b51419dbdc2012f34e1b
```
(gdb) frame 0
#0 0x7f3a36b4d5ac in update_dialog_vars_dbinfo (cell=0x7f3a23da0df8,
var=0x3433) at dlg_db_handler.c:652
652 SET_STR_VALUE(values+2, var->key);
(gdb) p *cell
$1 = {ref = 1, n
OK. I am updating the kamailio commit for the load tests.
Also, one of the kamailio instances running commit
2962852bf706692bdbc9b51419dbdc2012f34e1b at DigitalOcean crashed. If you want
access to it, just provide me the email address to where I should send the
access information.
---
Reply to
I am testing commit cecdd0817052046e634152d6e512165451b5f6d4 in our lab.
The crash happened again, but this time, it was when processing "200 OK":
```
(gdb) bt
#0 0x7f060c5ee5fc in update_dialog_vars_dbinfo (cell=0x7f05f919c150,
var=0x3433) at dlg_db_handler.c:652
#1 0x7f060c5ef449 in up
Well, I always do full build. So i did:
rm -fr kamailio
git clone git clone git://git.kamailio.org/kamailio kamailio
cd kamailio
git checkout cecdd0817052046e634152d6e512165451b5f6d4
Then, since cecdd0817052046e634152d6e512165451b5f6d4 is newer than
7f9194c416e171c2b723430eb98a55a52b7b5e38, I am s
27 matches
Mail list logo
