[tlug] Re: [RFC] Outline of the fast HTTP talk

[Edward Middleton <emiddleton@example.com>]

John Fremlin wrote:
> Edward Middleton <emiddleton@example.com> writes:
> 
>> Curt Sampson wrote:
>>> On 2008-11-03 18:56 +0000 (Mon), John Fremlin wrote:
>>>   
>>>> I get about 2k requests per second on one core of my laptop with the
>>>> following mongrel script....
>>>>
>>>> Are you in that ballpark? 
>>>>     
>>> Well, I've not benchmarked recently, but last time I checked, we were
>>> handling more than 500 requests per second on a 700 MHz Pentium II, so
>>> I'd say, yes we certainly are. Probably, given that on a multi-core
>>> machine we can distribute the load amongst multiple back-ends, we're
>>> faster. (Or does mongrel have the capability to use multiple
>>> processes?)
> 
> Mongrel is multithreaded
> 
>> My varnish server is Celeron 1.8GHz with 2G of ram.  I would be
>> interested in running any benchmarks you have against this server, for
>> comparison.
> 
> Feel free.
> 
> Run apachebench against this ruby script.
> 
> The test is for performance on one core, so if you have more use
> schedtool to set the CPU affinity (and run apachebench on another core).
> 
> schedtool -a 0 -e ruby mongrels.rb
> schedtool -a 1 -e ab -n 10000 -c100 http://127.0.0.1:3000/?name=TLUG

Running ab against the front page I got i.e.

# ab -n 10000 -c 100 http://blog.vortorus.net/


Server Software:        nginx/0.5.37
Server Hostname:        blog.vortorus.net
Server Port:            80

Document Path:          /
Document Length:        21179 bytes

Concurrency Level:      100
Time taken for tests:   14.922 seconds
Complete requests:      10000
Failed requests:        0
Write errors:           0
Total transferred:      215150000 bytes
HTML transferred:       211790000 bytes
Requests per second:    670.16 [#/sec] (mean)
Time per request:       149.218 [ms] (mean)
Time per request:       1.492 [ms] (mean, across all concurrent requests)
Transfer rate:          14080.58 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    1   1.8      0      22
Processing:    33  148  27.3    144     522
Waiting:       31  147  26.1    143     521
Total:         40  149  27.4    144     524

Percentage of the requests served within a certain time (ms)
  50%    144
  66%    151
  75%    155
  80%    159
  90%    169
  95%    177
  98%    202
  99%    261
 100%    524 (longest request)



using a static version of the page you gave I got.

# ab -n 10000 -c 100 http://blog.vortorus.net/tst.html


Server Software:        nginx/0.5.37
Server Hostname:        blog.vortorus.net
Server Port:            80

Document Path:          /tst.html
Document Length:        20 bytes

Concurrency Level:      100
Time taken for tests:   14.268 seconds
Complete requests:      10000
Failed requests:        0
Write errors:           0
Total transferred:      3180000 bytes
HTML transferred:       200000 bytes
Requests per second:    700.85 [#/sec] (mean)
Time per request:       142.684 [ms] (mean)
Time per request:       1.427 [ms] (mean, across all concurrent requests)
Transfer rate:          217.65 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    0   0.9      0      13
Processing:    31  142  35.3    137     432
Waiting:       30  141  35.2    137     432
Total:         37  142  35.2    137     433

Percentage of the requests served within a certain time (ms)
  50%    137
  66%    145
  75%    150
  80%    154
  90%    174
  95%    216
  98%    263
  99%    277
 100%    433 (longest request)