Hi,

I've tested your configuration with my current patch (not published yet), here 
is the results:

# ab -n 100000 -c 10 -k http://192.168.1.47/ ; sleep 15 ; echo ; ./stats
This is ApacheBench, Version 2.3 <$Revision: 655654 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking 192.168.1.47 (be patient)
Completed 10000 requests
Completed 20000 requests
Completed 30000 requests
Completed 40000 requests
Completed 50000 requests
Completed 60000 requests
Completed 70000 requests
Completed 80000 requests
Completed 90000 requests
Completed 100000 requests
Finished 100000 requests


Server Software:        Apache/2.2.17
Server Hostname:        192.168.1.47
Server Port:            80

Document Path:          /
Document Length:        12 bytes

Concurrency Level:      10
Time taken for tests:   3.664 seconds
Complete requests:      100000
Failed requests:        0
Write errors:           0
Keep-Alive requests:    99016
Total transferred:      33057374 bytes
HTML transferred:       1200024 bytes
Requests per second:    27291.42 [#/sec] (mean)
Time per request:       0.366 [ms] (mean)
Time per request:       0.037 [ms] (mean, across all concurrent requests)
Transfer rate:          8810.38 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    0   0.0      0       1
Processing:     0    0   0.3      0      19
Waiting:        0    0   0.3      0      19
Total:          0    0   0.3      0      19

Percentage of the requests served within a certain time (ms)
  50%      0
  66%      0
  75%      0
  80%      0
  90%      1
  95%      1
  98%      1
  99%      1
 100%     19 (longest request)

   5 nobody
   5 www
--------------------
  10 TOTAL

* Please note that nobody is the multiplexer and www the worker.

Please run the the following command:

ab -n 100000 -c 10 -k http://YOURHOST/ ; sleep 15; echo ; ./stats

Here is the stats source:

------------
#!/bin/sh

cd /tmp
ps -aux > tmp.txt
cat tmp.txt | grep httpd | cut -d" " -f1 | grep -v "^root$" > tmp2.txt
cat tmp2.txt | sort | uniq -c | sort -n
echo "--------------------"
echo -n "  "
echo -n `cat tmp.txt | grep -v "^root" | grep -c httpd`
echo " TOTAL"
rm tmp.txt
rm tmp2.txt
------------

Please post back the results of this commands.

--
Marcelo Coelho
marcelo at mco2.com.br




On Dec 30, 2010, at 6:21 AM, Taavi Sannik wrote:

> Hello again!
> 
> It seems that new patch does not always honor the process limits.
> 
> For example I have this in my configuration:
> <IfModule peruser.c>
>        ServerLimit 200
>        MaxClients 200
>        MinSpareProcessors 1
>        MinProcessors 5
>        MaxProcessors 150
>        MaxRequestsPerChild 10000
>        ExpireTimeout 30
>        IdleTimeout 10
>        MinMultiplexers 5
>        MaxMultiplexers 15
>        MultiplexerIdleTimeout 10
>        ProcessorWaitTimeout 10
> </IfModule>
> (I know the timeouts are very low, just don't want to wait for the children 
> to die)
> 
> So if I send a request spike to the server (10k requests for example) and 
> then wait for the idle timeout to kick in, then my server environment will 
> not always have 5 workers (which is my minimum limit). In fact many cases I 
> tested, my server environment has no workers at all.
> 
> The case actually gets worse as it sometimes works the same way with 
> multiplexers. I have been using this patch on my development server (which 
> has very low traffic) and sometimes if I just leave the server running and 
> check back later, then I see a dead server, because all the multiplexers have 
> died.
> 
> There are no errors or warnings about children segfaulting or anything 
> suggesting that server is killing the children by itself.
> 
> -- 
> Taavi Sannik
> DataCode OY
> 
> _______________________________________________
> Peruser mailing list
> [email protected]
> http://www.telana.com/mailman/listinfo/peruser
> 

_______________________________________________
Peruser mailing list
[email protected]
http://www.telana.com/mailman/listinfo/peruser

Reply via email to