-
Notifications
You must be signed in to change notification settings - Fork 3.5k
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
worker process exited on signal 11 #470
Comments
I made some progress with gdb: Program received signal SIGSEGV, Segmentation fault. |
which nginx version ? |
/usr/local/nginx/sbin/nginx -V |
Got the same shit on 1.7.5 |
I am using 1.7.5, everything seems OK! |
Same problem on 1.7.5 , OS: Centos 6 |
The problem seems to occur with random static cam like these: |
Just FYI. See also comments in #472. |
I have recompiled nginx with debug option to get some more info: 2014/09/19 12:31:13 [debug] 2503#0: posted event 0000000000000000 |
Please update the code from master. It should be fixed now. |
I rebuild the server with nginx 1.6.2 and git://github.com/arut/nginx-rtmp-module.git |
@arut Awesome! Just tested, works like a charm in both scenarios described above (HTTP only and with RTMP config). Thanks again a ton for fixing this so quickly. |
The config what I use for testing and, still get these errors:
|
Without the pull section I don't get these errors, but with the static cams the worker process stops approximately in every 10 min. |
@gmarcz what if you recompile with nginx 1.7.5? |
Strange because I got the same errors with 1.7.5 nginx version: nginx/1.7.5 2014/09/22 13:23:19 [alert] 24#0: worker process 3841 exited on signal 11 Program received signal SIGSEGV, Segmentation fault. |
@gmarcz please set up debug logging and post more lines from error.log |
@gmarcz Didn't wait long enough during my tests... I get the signal 11 error too after about 20min. |
These are the last few lines from the exiting process: 2014/09/22 15:43:00 [debug] 353#0: *15 nhandlers: 5 |
Same problem in 1.7.5 |
Please try Thanks everyone in this thread for the reports. I'm waiting for feedback. |
I've tested your fix playing 2 random streams from the configuration for 20min+; videos are still playing, no entries in error log. I'll leave them running for longer, but it looks good to me. My compile configuration (with branch static-relay-cleanup): nginx version: nginx/1.7.5 |
Update: 1h40m and the videos are still streaming with above configuration. I guess this works. |
@arut thank you for your effort, I will test it with 1.6.2 and let you know if I found anything. |
Server is running about 2h 49m seems to be stable for now. |
We have the same problem with pulling, but our pulls aren't static. 2014/09/24 04:13:59 [alert] 15065#0: worker process 29253 exited on signal 11 nginx 1.7.5, latest nginx-rtmp version. The static-relay-cleanup didn't help. It's hard to debug because of heavy traffic. Will the debug log be enough, or I should also run gdb? |
@developer222 yes, please report gdb backtrace |
Hello, We have a similar problem with dying worker on a server with high number of active streams (1000-1500). The traffic isn't so big, about 300 Mbps to 1.5 Gbps. I've recompiled static-cleanup-branch and nginx 1.6.2 with debug enabled and here is what we have in gdb output:Core was generated by `nginx: worker process '. ccf = 0x2236258nginx -V nginx config: user nginx; for debug only:worker_rlimit_core 2000M; error_log /var/log/nginx/error.log warn; events { rtmp {
... Do you have any insights how to mitigate this issue? Thanks! |
@arut , sorry for the long delay. Here's the gdb output with backtrace. What else may I provide to clarify the problem? Thanks. gdb /usr/sbin/nginx.debug core warning: Can't read pathname for load map: Input/output error. |
Hi, I have came same error in nginx 1.10 version. Can you explain where is exact problem. |
Still there today on Nginx 1.11.1 |
try add |
Hi All!
From time to time I see this error in error.log
2014/09/15 11:34:24 [alert] 27#0: worker process 28 exited on signal 11
2014/09/15 11:44:28 [alert] 27#0: worker process 21539 exited on signal 11
2014/09/15 12:34:48 [alert] 27#0: worker process 9652 exited on signal 11
2014/09/15 12:44:53 [alert] 27#0: worker process 15638 exited on signal 11
2014/09/15 12:49:20 [alert] 27#0: worker process 32588 exited on signal 11
2014/09/15 12:53:49 [alert] 27#0: worker process 4974 exited on signal 11
2014/09/15 13:12:06 [alert] 27#0: worker process 10578 exited on signal 11
2014/09/15 13:28:17 [alert] 27#0: worker process 2807 exited on signal 11
2014/09/15 13:35:06 [alert] 27#0: worker process 30753 exited on signal 11
I don't have to much experience to debug such problem, what I tried was to attache strace to the worker process, the last few lines from the log:
kill(26701, SIGKILL) = 0
--- SIGCHLD {si_signo=SIGCHLD, si_code=CLD_KILLED, si_pid=26701, si_status=SIGKILL, si_utime=1, si_stime=0} ---
gettimeofday({1410787697, 694141}, NULL) = 0
wait4(-1, [{WIFSIGNALED(s) && WTERMSIG(s) == SIGKILL}], WNOHANG, NULL) = 26701
write(4, "2014/09/15 13:28:17 [alert] 2807"..., 77) = 77
wait4(-1, 0x7fff044965fc, WNOHANG, NULL) = 0
rt_sigreturn() = 0
sendto(63, "\2\0\0\0\0\0\6\4\0\0\0\0\0\1\0\0\0\1", 18, 0, NULL, 0) = -1 EPIPE (Broken pipe)
--- SIGPIPE {si_signo=SIGPIPE, si_code=SI_USER, si_pid=2807, si_uid=65534} ---
close(16) = 0
close(63) = 0
--- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x10} ---
+++ killed by SIGSEGV +++
Can someone help to identify the root case of this problem ?
Thank you for your time.
Regards
Gabor
The text was updated successfully, but these errors were encountered: