[Home]

Summary:ASTERISK-14300: After 20 mins with heaviy traffic it blocks
Reporter:Private Name (falves11)Labels:
Date Opened:2009-06-10 15:21:30Date Closed:2009-06-15 12:23:33
Priority:BlockerRegression?No
Status:Closed/CompleteComponents:Channels/chan_sip/General
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:
Description: utils.c:1074 ast_wait_for_output: Timed out trying to write

I am suddenly out of business. Is it a version issue or something is wromg with my box?

****** ADDITIONAL INFORMATION ******

do know what happens, it started today. this is my ulimit
core file size          (blocks, -c) unlimited
data seg size           (kbytes, -d) unlimited
scheduling priority             (-e) 0
file size               (blocks, -f) unlimited
pending signals                 (-i) 400000
max locked memory       (kbytes, -l) 32
max memory size         (kbytes, -m) unlimited
open files                      (-n) 400000
pipe size            (512 bytes, -p) 8
POSIX message queues     (bytes, -q) 819200
real-time priority              (-r) 0
stack size              (kbytes, -s) 10240
cpu time               (seconds, -t) unlimited
max user processes              (-u) 1056768
virtual memory          (kbytes, -v) unlimited
file locks                      (-x) unlimited

my rtp.conf
[general]
rtpstart=5000
rtpend=62767
rtpchecksums=no
dtmftimeout=3000

I am not using udptl.conf, but its range is 1500-4999

Comments:By: Private Name (falves11) 2009-06-11 19:46:55

if I do this at the Linux command prompt:
lsof -i | wc -l

The number keeps groing even if we keep the same amount calls, and calls per seconds, flowing through. That's why it blows up after a while. everything started, I am guessing, when I added this to file the sip.conf
t1min=500
session-timers=accept
session-expires=3600
session-minse=90
session-refresher=uac

Which we need because our carriers keep the calls hanging.

Any ideas?
By: Private Name (falves11) 2009-06-11 19:56:11

It does not seem to be related. The growth still goes on, however, slower.
By: Private Name (falves11) 2009-06-11 20:24:15

It is related and I can prove it. The amount of handles declines when we are not using Session Timers and the amount of calls goes down. So the culprit is Session Timers, it has a handle leak.
By: Private Name (falves11) 2009-06-12 09:43:33

It is not related. I disabled Session Timers and UDPTL (T38) support in my sip.conf, but the handle cpunt keeps growing. It is now in 15600 handles with only 545 calls open, and I am using directrtpsetup=yes, so no media goes through me.


By: Private Name (falves11) 2009-06-12 13:54:59

It actually crashes
[Jun 12 14:47:52] WARNING[23793]: channel.c:2582 __ast_read: read() failed: Bad file descriptor
[Jun 12 14:47:52] ERROR[23793]: astobj2.c:110 INTERNAL_OBJ: user_data is NULL
balancer*CLI>

maybe there us another bug open?
By: Private Name (falves11) 2009-06-15 09:07:42

Version 1.4 does not have any handle leak. It shows a steady correlation of about 4 handles per open call, while any of the 1.6.X versions this relationship is a lot bigger, maybe 10. I am going back to 1.4, but we should look into this and fix it.
I can show the developers both cases, with the same calls, and you can see the huge difference.


By: Leif Madsen (lmadsen) 2009-06-15 12:23:32

This is a duplicate of an existing bug. Issue 13623 mentions a work around.