Closed kamil-michalak closed 10 years ago
Do you have a core dump? If yes, send it to me, and you nginx executable too.
What is the exact version of push stream module you are using?
First, I set limit for core dump to low (512MB). Now i changed it to 3GB and still waiting (12-24 hours to next segfault). I use dotdeb package (1.4.2) on debian wheezy 64bit http://packages.dotdeb.org/dists/wheezy/nginx/source/ it's contains Version 0.3.5
Hi, can you check if this happens with the new version 0.4.0 3d3a204177d3a7ab8a2858e04e792a6d11bf133f ?
Hi Wandenberg,
No. but I'll try and let you know
Nov 12 07:09:55 xx kernel: [5974241.116469] nginx[51257]: segfault at 0 ip 00000000004d9bf7 sp 00007fffd00aaa80 error 6 in nginx[400000+b95000] Nov 12 07:09:55 x kernel: [5974241.140292] nginx[49838]: segfault at 0 ip 00000000004d9bf7 sp 00007fffd00aaad0 error 6 in nginx[400000+b95000]
Can you provide a core dump and your binary? Which modules are you using? Send me your configuration and how to reproduce the problem.
Hi @kilgor,
please, try the code on 0.4.1-dev branch, specially the commit 51356e5621d6c5491d8556e315a03e6a6317e585 solve this problem.
This version works without a problem since two weeks
this fix was merged to master
I have some problems ;)
(gdb) bt
0 ngx_http_push_stream_worker_subscriber_cleanup_locked (worker_subscriber=0x27deb50) at /usr/src/nginx/source/nginx-1.4.2/debian/modules/nginx-push-stream-module/src/ngx_http_push_stream_module_utils.c:1172
1 ngx_http_push_stream_cleanup_request_context (r=<error reading variable: Cannot access memory at address 0x7fff76f96398>)
If you neeed more information i can help you