Login | Register For Free | Help
Search for: (Advanced)

Mailing List Archive: Apache: Bugs
[Bug 52914] New: On keep-alive on, after some time huge process load
 

Index | Next | Previous | View Flat


bugzilla at apache

Mar 15, 2012, 1:28 AM


Views: 109
Permalink
[Bug 52914] New: On keep-alive on, after some time huge process load

https://issues.apache.org/bugzilla/show_bug.cgi?id=52914

Bug #: 52914
Summary: On keep-alive on, after some time huge process load
Product: Apache httpd-2
Version: 2.4.1
Platform: PC
OS/Version: Linux
Status: NEW
Severity: critical
Priority: P2
Component: Core
AssignedTo: bugs [at] httpd
ReportedBy: dohardgopro [at] gmail
Classification: Unclassified


Created attachment 28470
--> https://issues.apache.org/bugzilla/attachment.cgi?id=28470
patch

System:
Linux 3.0.0-1-amd64 #1 SMP Sat Aug 27 16:21:11 UTC 2011 x86_64 GNU/Linux
Apache version
Server version: Apache/2.2.22 (Debian)

But I sure that at 2.4.1 the same issue

After turning set next options:
KeepAlive On
MaxKeepAliveRequests 5
KeepAliveTimeout 1

After some time (at last iteration about 2 minutes) some of apache workers eat
60-100% CPU, while must 0-2% CPU (without keep-alive - so)

Gdb:
(gdb) bt
#0 0x00007f61b5bb0cad in apr_table_setn () from /usr/lib/libapr-1.so.0
#1 0x00007f61af75eaca in reqtimeout_filter (f=0x7f61b6275b40,
bb=0x7f61b6266170, mode=<optimized out>, block=<optimized out>,
readbytes=<optimized out>) at mod_reqtimeout.c:309
#2 0x00007f61b3146322 in dumpio_input_filter (f=0x7f61b6275a68,
bb=0x7f61b6266170, mode=AP_MODE_READBYTES, block=APR_BLOCK_READ,
readbytes=6440) at mod_dumpio.c:120
#3 0x00007f61b64b0161 in ap_http_filter (f=0x7f61b6263b10, b=0x7f61b6266170,
mode=AP_MODE_READBYTES, block=<optimized out>, readbytes=6440) at
http_filters.c:525
#4 0x00007f61b64b1914 in ap_discard_request_body (r=0x7f61b6261758) at
http_filters.c:1366
#5 0x00007f61b6491349 in ap_finalize_request_protocol (r=0x7f61b6261758) at
protocol.c:1168
#6 0x00007f61af552a05 in handler_redirect (r=0x7f61b62620a0) at
mod_rewrite.c:4860
#7 0x00007f61b649f668 in ap_run_handler (r=0x7f61b62620a0) at config.c:159
#8 0x00007f61b649fade in ap_invoke_handler (r=0x7f61b62620a0) at config.c:377
#9 0x00007f61b64af6b0 in ap_process_request (r=0x7f61b62620a0) at
http_request.c:282
#10 0x00007f61b64ac4f8 in ap_process_http_connection (c=0x7f61b6275290) at
http_core.c:190
#11 0x00007f61b64a60e8 in ap_run_process_connection (c=0x7f61b6275290) at
connection.c:43
#12 0x00007f61b64b42c0 in child_main (child_num_arg=<optimized out>) at
prefork.c:667
#13 0x00007f61b64b4a2a in make_child (slot=6, s=0x7f61b64407f8) at
prefork.c:768
#14 make_child (s=0x7f61b64407f8, slot=6) at prefork.c:696
#15 0x00007f61b64b55df in perform_idle_server_maintenance (p=<optimized out>)
at prefork.c:903
#16 ap_mpm_run (_pconf=<optimized out>, plog=<optimized out>, s=<optimized
out>) at prefork.c:1107
#17 0x00007f61b648a6a4 in main (argc=3, argv=0x7fff67bcc5e8) at main.c:755
(gdb) quit
A debugging session is active.

Inferior 1 [process 11587] will be detached.

Quit anyway? (y or n) EOF [assumed Y]
Detaching from program: /usr/lib/apache2/mpm-prefork/apache2, process 11587
root@:/work1/azat/apache_high_percents/cores# ps u 11587
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
www-data 11587 59.0 0.0 381148 14968 ? R 14:33 0:42
/usr/sbin/apache2 -k start
root@:/work1/azat/apache_high_percents/cores# ps u 11587
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
www-data 11587 60.0 0.0 381148 14968 ? R 14:33 0:45
/usr/sbin/apache2 -k start
root@:/work1/azat/apache_high_percents/cores# ps u 11587
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
www-data 11587 60.4 0.0 381148 14968 ? R 14:33 0:45
/usr/sbin/apache2 -k start
root@:/work1/azat/apache_high_percents/cores# ps u 11587
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
www-data 11587 61.5 0.0 381148 14968 ? R 14:33 0:46
/usr/sbin/apache2 -k start

(gdb) bt
#0 0x00007f61b64a980c in ap_get_brigade (next=0x7f61b6263b10,
bb=0x7f61b6266170, mode=AP_MODE_READBYTES, block=APR_BLOCK_READ,
readbytes=8192) at util_filter.c:489
#1 0x00007f61b64b1914 in ap_discard_request_body (r=0x7f61b6261758) at
http_filters.c:1366
#2 0x00007f61b6491349 in ap_finalize_request_protocol (r=0x7f61b6261758) at
protocol.c:1168
#3 0x00007f61af552a05 in handler_redirect (r=0x7f61b62620a0) at
mod_rewrite.c:4860
#4 0x00007f61b649f668 in ap_run_handler (r=0x7f61b62620a0) at config.c:159
#5 0x00007f61b649fade in ap_invoke_handler (r=0x7f61b62620a0) at config.c:377
#6 0x00007f61b64af6b0 in ap_process_request (r=0x7f61b62620a0) at
http_request.c:282
#7 0x00007f61b64ac4f8 in ap_process_http_connection (c=0x7f61b6275290) at
http_core.c:190
#8 0x00007f61b64a60e8 in ap_run_process_connection (c=0x7f61b6275290) at
connection.c:43
#9 0x00007f61b64b42c0 in child_main (child_num_arg=<optimized out>) at
prefork.c:667
#10 0x00007f61b64b4a2a in make_child (slot=6, s=0x7f61b64407f8) at
prefork.c:768
#11 make_child (s=0x7f61b64407f8, slot=6) at prefork.c:696
#12 0x00007f61b64b55df in perform_idle_server_maintenance (p=<optimized out>)
at prefork.c:903
#13 ap_mpm_run (_pconf=<optimized out>, plog=<optimized out>, s=<optimized
out>) at prefork.c:1107
#14 0x00007f61b648a6a4 in main (argc=3, argv=0x7fff67bcc5e8) at main.c:755
(gdb) quit
A debugging session is active.

Inferior 1 [process 11587] will be detached.

Quit anyway? (y or n) EOF [assumed Y]
Detaching from program: /usr/lib/apache2/mpm-prefork/apache2, process 11587
root@:/work1/azat/apache_high_percents/cores# ps u 11587
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
www-data 11587 71.9 0.0 381148 14968 ? R 14:33 2:11
/usr/sbin/apache2 -k start

Server-status
Srv PID Acc M CPU SS Req Conn Child Slot Client VHost Request
6-0 11587 0/2/17 W 0.03 194 0 0.0 0.00 0.00 {CLIENT_IP} {SERVER_HOST} POST
{URL} HTTP/1.1

While here CPU = 0.03

I also think that my patch can fix this problem
There I write verbose comment

--
Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe [at] httpd
For additional commands, e-mail: bugs-help [at] httpd

Subject User Time
[Bug 52914] New: On keep-alive on, after some time huge process load bugzilla at apache Mar 15, 2012, 1:28 AM

  Index | Next | Previous | View Flat
 
 


Interested in having your list archived? Contact Gossamer Threads
 
  Web Applications & Managed Hosting Powered by Gossamer Threads Inc.