[Freeswitch-dev] Strange bug - 2
Oleg Khovayko
khovayko at gmail.com
Thu Sep 16 07:29:20 PDT 2010
Rupa Schomaker wrote:
> I should note that you are looking for identical stack traces (up to a
> certain depth) that are not blocked (eg on a mutex). If there is a
> runaway thread (100% cpu implies this) then it should be visible in
> the stack traces.
>
Excuse me, I don't know, how to see stack trace in the running process,
without stopping it.
> I assume you tried running with loglevel 7 and there is nothing useful
> in the logs?
This time - no. But, when analogous situation happens ~1 month ago, I
tried to use loglevel 7, ans siptrace on - did not found any suspicious
message.
Prev time, it was happeining, also keep for ~2 days, and wanished.
Reboot, reinstall binaries, was not help.
Looks like deadlock with some files access.
This time, I have fixed (or, maybe problem vanished itself again) by
following:
I disabled load modules mod_portaudio and mod_dingaling.
Thereafter, program eats ~4%CPU (not 60%), memory does not leaks, and FS
works.
But, prev. time, program consumed ~2%. I think, problem persist, but
just hidden.
Also, when I run fs manually, and use "shutdown", then shutdown log
freeze on something like "Wait for pending SQL transaction".
I tried to clear ./db directory, and restart FS - does not help...
>
> On Thu, Sep 16, 2010 at 6:39 AM, Rupa Schomaker <rupa at rupa.com
> <mailto:rupa at rupa.com>> wrote:
>
> That core isn't going to help much, it is probably due to running
> out of memory due to whatever is the "real" problem.
>
> 1) Review the "Reporting Bugs" page on the wiki.
>
OK, I will see, thank you.
>
> 2) After getting FS started and exhibiting the behavior, get a few
> core dumps separated by time (10s apart?). The procedure should
> be on the wiki. Use gcore.
>
I will do, when problem will raised again.
>
> 3) What platform is this? 32b or 64b?
>
Intel-32, celeron-300.
>
>
>
> On Wed, Sep 15, 2010 at 7:05 PM, Oleg Khovayko <khovayko at gmail.com
> <mailto:khovayko at gmail.com>> wrote:
>
> As I wrote before, my FS has strange behaviour:
> When it starts, it creates all threads, and go to command prompt.
> But, CPU usage so high, and program eat more and more memory.
>
> When I invoke "shutdown", ito goes to shutdown, and after destroy
> gateway, freeze for ~2 minutes. Thereafter, crashes to core.
>
> GDB output following.
>
> PS: This is today's GIT build:
>
>
> 2010-09-15 19:57:45.229882 [NOTICE] sofia_glue.c:4350 deleted
> gateway
> example.com <http://example.com>
> 2010-09-15 19:57:45.229882 [NOTICE] sofia_glue.c:4350 deleted
> gateway
> voipdiscount.com <http://voipdiscount.com>
> 2010-09-15 19:57:45.229882 [NOTICE] sofia_glue.c:4350 deleted
> gateway
> sipsorcery.com <http://sipsorcery.com>
> 2010-09-15 19:57:45.229882 [NOTICE] sofia_glue.c:4350 deleted
> gateway
> gizmo5.com <http://gizmo5.com>
> 2010-09-15 19:59:54.935939 [NOTICE] switch_loadable_module.c:572
> Deleting Application 'hash'
> 2010-09-15 19:59:54.944150 [NOTICE] switch_loadable_module.c:599
> Deleting API Function 'hash'
> 2010-09-15 19:59:54.948691 [NOTICE] switch_loadable_module.c:599
> Deleting API Function 'hash_dump'
> 2010-09-15 19:59:54.951357 [NOTICE] switch_loadable_module.c:599
> Deleting API Function 'hash_remote'
> 2010-09-15 19:59:54.954504 [NOTICE] switch_loadable_module.c:803
> Deleting Limit interface 'hash'
> 2010-09-15 19:59:54.958190 [CONSOLE] switch_loadable_module.c:1396
> Stopping: mod_hash
> 2010-09-15 19:59:54.960612 [NOTICE] switch_loadable_module.c:557
> Deleting Timer 'soft'
> 2010-09-15 19:59:54.960612 [CONSOLE] switch_loadable_module.c:1396
> Stopping: CORE_SOFTTIMER_MODULE
> 2010-09-15 19:59:54.965641 [CONSOLE] switch_time.c:852 Soft
> timer thread
> exiting.
> 2010-09-15 19:59:54.966260 [NOTICE]
> switch_loadable_module.c:99 Thread
> ended for CORE_SOFTTIMER_MODULE
> 2010-09-15 19:59:54.977525 [NOTICE] switch_event.c:1500 Event
> Binding
> deleted for CORE_SOFTTIMER_MODULE:RELOADXML
> 2010-09-15 19:59:54.980836 [NOTICE] switch_loadable_module.c:491
> Deleting Endpoint 'dingaling'
> 2010-09-15 19:59:54.983836 [NOTICE] switch_loadable_module.c:599
> Deleting API Function 'dl_debug'
> [New Thread 0x28802030 (LWP 100077)]
>
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 0x28802030 (LWP 100077)]
> sqlite3HashInsert (pH=0x0, pKey=0x2959cd78, nKey=15,
> data=0x28c41080) at
> ./src/hash.c:347
> 347 xHash = hashFunction(pH->keyClass);
> (gdb) where
> #0 sqlite3HashInsert (pH=0x0, pKey=0x2959cd78, nKey=15,
> data=0x28c41080) at ./src/hash.c:347
> #1 0x280e972f in switch_core_hash_insert (hash=0x0,
> key=0x2959cd78
> "173.79.240.220", data=0x28c41080) at src/switch_core_hash.c:80
> #2 0x28eb0969 in sofia_glue_add_profile (key=0x2959cd78
> "173.79.240.220", profile=0x28c41080) at sofia_glue.c:4246
> #3 0x28ed4d50 in sofia_reg_handle_sip_i_register (nua=0x2882a240,
> profile=0x28c41080, nh=0x2b5f2de0, sofia_private=0x0,
> sip=0x2b5df89c,
> tags=0x2b6033bc) at sofia_reg.c:1442
> #4 0x28eaa5a5 in sofia_event_callback (event=nua_i_register,
> status=100, phrase=0x2b6033c4 "Trying", nua=0x2882a240,
> profile=0x28c41080,
> nh=0x2b5f2de0, sofia_private=0x0, sip=0x2b5df89c,
> tags=0x2b6033bc)
> at sofia.c:912
> #5 0x28f1e866 in nua_application_event (dummy=0x0,
> sumsg=0xbf354cd8,
> ee=0x2b6033a4) at nua_stack.c:393
> #6 0x28f6ec3c in su_base_port_execute_msgs (queue=0x0) at
> su_base_port.c:280
> #7 0x28f6f4a8 in su_base_port_run (self=0x28806c50) at
> su_base_port.c:335
> #8 0x28f6a6a1 in su_root_run (self=0x28c46200) at su_port.h:326
> #9 0x28ea1a59 in sofia_profile_thread_run (thread=0x28c41d68,
> obj=0x28c41080) at sofia.c:1571
> #10 0x2817b436 in dummy_worker (opaque=0x28c41d68) at
> threadproc/unix/thread.c:138
> #11 0x282766ff in pthread_getprio () from /lib/libthr.so.3
> #12 0x00000000 in ?? ()
> (gdb)
>
>
>
>
> _______________________________________________
> FreeSWITCH-dev mailing list
> FreeSWITCH-dev at lists.freeswitch.org
> <mailto:FreeSWITCH-dev at lists.freeswitch.org>
> http://lists.freeswitch.org/mailman/listinfo/freeswitch-dev
> UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-dev
> http://www.freeswitch.org
>
>
>
>
> --
> -Rupa
>
>
>
>
> --
> -Rupa
>
>
> _______________________________________________
> FreeSWITCH-dev mailing list
> FreeSWITCH-dev at lists.freeswitch.org
> http://lists.freeswitch.org/mailman/listinfo/freeswitch-dev
> UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-dev
> http://www.freeswitch.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freeswitch.org/pipermail/freeswitch-dev/attachments/20100916/8515e466/attachment-0001.html
More information about the FreeSWITCH-dev
mailing list