lighttpd forum XCache > Has anyone benchmarked xcache yet ?

Posted by Malte Geierhos (Guest)
on 13.05.2006 17:39
It would be nice to see some performance comparison between eA, APC, 
ZendOptimizationSuite and xcache.

Has anyone done some tests until now ?
Posted by moo XCache
on 19.05.2006 05:01
Malte Geierhos wrote:
> It would be nice to see some performance comparison between eA, APC, 
> ZendOptimizationSuite and xcache.
> 
> Has anyone done some tests until now ?

xcache is young online so not too much user using it, and it seems no 
one is doing the benchmark.
but what i can tell u so far is:
i have done some benchmark sometimes ago, most non-stupid-non-dirty 
opcode cacher, if it's really doing opcode cacher, can boost php speed 
up about 2 times faster, and their benchmark is almost equal, with minor 
--- something like 0.01 to 0.1 percent --- difference.
and you can always see the official author claim their cacher is more 
faster than anyone's else.
Posted by jfbus (Guest)
on 02.06.2006 14:42
moo moo wrote:
> Malte Geierhos wrote:
>> It would be nice to see some performance comparison between eA, APC, 
>> ZendOptimizationSuite and xcache.
>> 
>> Has anyone done some tests until now ?
> 
> xcache is young online so not too much user using it, and it seems no 
> one is doing the benchmark.
> but what i can tell u so far is:
> i have done some benchmark sometimes ago, most non-stupid-non-dirty 
> opcode cacher, if it's really doing opcode cacher, can boost php speed 
> up about 2 times faster, and their benchmark is almost equal, with minor 
> --- something like 0.01 to 0.1 percent --- difference.
> and you can always see the official author claim their cacher is more 
> faster than anyone's else.

+1.

I'm trying to benchmark it. Already did it with apc & eAccelerator.

http://www.ipersec.com/index.php?q=en/bench_ea_vs_apc

The benchmark is recent, but I have some work to update it...
Posted by diefans (Guest)
on 24.01.2008 10:18
Posted by Jason Read (Guest)
on 27.02.2008 19:31
Malte Geierhos wrote:
> It would be nice to see some performance comparison between eA, APC, 
> ZendOptimizationSuite and xcache.
> 
> Has anyone done some tests until now ?

I've tested using an application containing about 50  php source files 
total of 78Mb cached. Compared against eAccelerator, xCache provides 
about the same performance increase. pecl-apc was about 20% slower.
Posted by Ruben Ortiz
on 13.03.2008 08:48
Jason Read wrote:
> Malte Geierhos wrote:
>> It would be nice to see some performance comparison between eA, APC, 
>> ZendOptimizationSuite and xcache.
>> 
>> Has anyone done some tests until now ?
> 
> I've tested using an application containing about 50  php source files 
> total of 78Mb cached. Compared against eAccelerator, xCache provides 
> about the same performance increase. pecl-apc was about 20% slower.

Excuse me by newbie question but.... How doy you do the test? What kind
of programs did you used? I have no ways or I don't know the properly
software to count and test the increase of perfomance.

If you give me some steps of do bencmark test I'll write a text in my 
blog speaking about XCache and the difference of load with APC/EA. 
Promise.

And about the test or benchamark, I tested the Xcache in a high server 
load and works perfectly. The day before we running with APC and we had 
a lot of troubles: PHP ACCESS VIOLATION, slowness, etc.

Bye!
Posted by moo XCache
on 14.03.2008 09:21
> Excuse me by newbie question but.... How doy you do the test? What kind
> of programs did you used? I have no ways or I don't know the properly
> software to count and test the increase of perfomance.
> 
> If you give me some steps of do bencmark test I'll write a text in my 
> blog speaking about XCache and the difference of load with APC/EA. 
> Promise.
just run some php app (vbulletin, phpmyadmin etc) with XCache enabled 
first, and then apc/ea etc. you gotta know how to configure each of them 
and tune it to be as fast as possible while keeping fair (like 
providing/cutting same set of function)
to do the test, run "ab" ("apache bench" tool from apache) or anything 
else that do mass request to webserver and show how fast the requests is 
done
the faster the better
it's easy for newbie to make mistake and produce wrong benchmark result 
however

> 
> And about the test or benchamark, I tested the Xcache in a high server 
> load and works perfectly. The day before we running with APC and we had 
> a lot of troubles: PHP ACCESS VIOLATION, slowness, etc.
it's big difference :)
but you may want some apc/ea guys help your troubleshooting
> 
> Bye!