here we are...
I also had the opportunity to refresh my kickstart install with the
final 6.0 and so scratch the beta2, so I got results also for that.

I used for client an external rh el 5.5 server where I have
httpd-2.2.3-43.el5_5.3
and
ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0

command line from client:
ab -n 100000 -c 5 http://10.4.5.xx/test.html

So the target servers are
Dell blades M610 with 48GB of ram and 2-quadcore sockets and HT
enabled (so OS sees 16 cpus)
test.html put on internal disk

RHEL 5.5 up2date (httpd-2.2.3-43.el5_5.3)
Requests per second:    8686.29 [#/sec] (mean)

RH EL 6.beta2 + kernel
Requests per second:    7724.68 [#/sec] (mean)

RH EL 6 (httpd-2.2.15-5.el6.x86_64)
Requests per second:    9007.28 [#/sec] (mean)


Note that running test from self gave results, in my opinion not meaningful:
RH EL 5.5 from self
Requests per second:    21508.65 [#/sec] (mean)

RH EL 6beta2 from self
Requests per second:    10117.01 [#/sec] (mean)

This reflects your initial results, but doesn't match the benchmark
results above that are run:
1) from external server
and
2) using same benchmark version

RH EL 6 final from self
Requests per second:    11983.33 [#/sec] (mean)

BTW: I noticed that on both rh el 6beta2 and rh el 6 final, after the
end of the bench, for about 1-2 minutes I keep having many and many of
the connections opened during the bench itself, in this state:
...
tcp        0      0 10.4.5.39:80                10.4.5.119:42447
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:32195
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:31138
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:45923
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:24058
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:42350
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:50975
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:36206
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:16189
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:45843
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:52303
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:49031
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:33901
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:48677
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:28692
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:36793
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:62740
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:62089
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:42205
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:47999
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:40877
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:20921
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:18546
     TIME_WAIT
tcp        0      0 10.4.5.39:80                10.4.5.119:40200
     TIME_WAIT
...

After about 2 minutes they are cleaned, but in rh el 5.5 it seems this
is faster...

Cheers,
Gianluca

_______________________________________________
rhelv6-beta-list mailing list
rhelv6-beta-list@redhat.com
https://www.redhat.com/mailman/listinfo/rhelv6-beta-list

Reply via email to