We just got a new box in and I did some benchmarks on it just to see how it
worked. It's an Intel i5 (dual core, hyperthreading enabled, so it appears to
have four cores to the OS) running 64-bit Ubuntu 9.10. Each core reports
something like this from /proc/cpuinfo:
processor : 3
vendor_id : GenuineIntel
cpu family : 6
model : 30
model name : Intel(R) Core(TM) i5 CPU 750 @ 2.67GHz
stepping : 5
cpu MHz : 1197.000
cache size : 8192 KB
physical id : 0
siblings : 4
core id : 3
cpu cores : 4
apicid : 6
initial apicid : 6
fpu : yes
fpu_exception : yes
cpuid level : 11
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat
pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm
constant_tsc arch_perfmon pebs bts rep_good xtopology tsc_reliable nonstop_tsc
pni dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2
popcnt lahf_lm ida tpr_shadow vnmi flexpriority ept vpid
bogomips : 5333.41
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
For the static file rendered at / we get this from ab -n1001 -c15:
Server Software: Apache/2.2.12
Server Hostname: 127.0.0.1
Server Port: 80
Document Path: /
Document Length: 177 bytes
Concurrency Level: 15
Time taken for tests: 0.088 seconds
Complete requests: 1001
Failed requests: 0
Write errors: 0
Total transferred: 452904 bytes
HTML transferred: 177354 bytes
Requests per second: 11339.31 [#/sec] (mean)
Time per request: 1.323 [ms] (mean)
Time per request: 0.088 [ms] (mean, across all concurrent requests)
Transfer rate: 5010.24 [Kbytes/sec] received
That would be somewhere around 11K requests per second or so.
For the followng "bare" WSGI app:
def application(environ, start_response):
start_response('200 OK', [('Content-Type', 'text/plain'),
('Content-Length', '5')])
return ['hello']
ab -n1001 -c15 returns:
Server Software: Apache/2.2.12
Server Hostname: 127.0.0.1
Server Port: 80
Document Path: /bareapp
Document Length: 5 bytes
Concurrency Level: 15
Time taken for tests: 0.090 seconds
Complete requests: 1001
Failed requests: 0
Write errors: 0
Total transferred: 180720 bytes
HTML transferred: 5020 bytes
Requests per second: 11129.89 [#/sec] (mean)
Time per request: 1.348 [ms] (mean)
Time per request: 0.090 [ms] (mean, across all concurrent requests)
Transfer rate: 1962.29 [Kbytes/sec] received
That's about 11K requests per second too.
For the "helloworld" program at http://svn.repoze.org/whatsitdoing/bfg using
mod_wsgi 1.2a2 under Python 2.6 configured using mod_wsgi 2.5 in daemon mode
with 4 processes and 15 threads, we get the following from ab -n1001 -c15:
Server Software: Apache/2.2.12
Server Hostname: 127.0.0.1
Server Port: 80
Document Path: /helloworld
Document Length: 5 bytes
Concurrency Level: 15
Time taken for tests: 0.111 seconds
Complete requests: 1001
Failed requests: 0
Write errors: 0
Total transferred: 179537 bytes
HTML transferred: 5015 bytes
Requests per second: 9021.27 [#/sec] (mean)
Time per request: 1.663 [ms] (mean)
Time per request: 0.111 [ms] (mean, across all concurrent requests)
Transfer rate: 1580.11 [Kbytes/sec] received
Using "siege" (http://freshmeat.net/projects/siege/) for the BFG app we get
similar results (1000 reps, 20 concurrent users in benchmark mode):
** SIEGE 2.68
** Preparing 20 concurrent users for battle.
The server is now under siege.. done.
Transactions: 20000 hits
Availability: 100.00 %
Elapsed time: 2.48 secs
Data transferred: 0.48 MB
Response time: 0.00 secs
Transaction rate: 8064.52 trans/sec
Throughput: 0.19 MB/sec
Concurrency: 19.61
Successful transactions: 20000
Failed transactions: 0
Longest transaction: 0.02
Shortest transaction: 0.00
That would be somewhere around 8K - 9K requests per second, depending on whom
you believe.
That's pretty snappy hardware for not much money. I don't think we have much
excuse for slow web apps anymore.
- C
_______________________________________________
Repoze-dev mailing list
Repoze-dev@lists.repoze.org
http://lists.repoze.org/listinfo/repoze-dev
