[Koha-devel] Benchmarking Koha

Chris Cormack chrisc at catalyst.net.nz
Fri Dec 2 18:56:32 CET 2011


* Dobrica Pavlinusic (dpavlin at gmail.com) wrote:
> On Fri, Dec 02, 2011 at 05:46:35PM +0100, Paul Poulain wrote:
> > My first surprise (and I did the tests 3 times to be sure I wasn't
> > making a mistake !) is that you almost get no speed improvement with
> > memcache ON.
> 
> I have seen similar behaviour when I benchmarked Koha about a year
> ago[1]
> 
> In the end, I opted to test with plain Memoize as opposed to
> memcache-backed one. My assumption was that TCP overhead eats away all
> improvements we might get from memcache (or to put it nicely, mysql and
> zebra are not *that* slow :-)
> 
> 1: http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=7177
> 2: http://git.rot13.org/?p=koha.git;a=commitdiff;h=b2155fc483f09b34c4a6ba92256f2732152bb1d5;hp=daadc5bc8f24e1bf2c1e8d958d410408d1cccc47
> 
Memcached is for scalability not speed. It will however give you
speed in easy serialised and deserialised objects, that are costly to
create.

But scalability is the main thing, you will notice better throughput.
And it will take load off your database server, which when you site is
very busy will provide speed.

CHris

--
Chris Cormack
Catalyst IT Ltd.
+64 4 803 2238
PO Box 11-053, Manners St, Wellington 6142, New Zealand
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: </pipermail/koha-devel/attachments/20111203/ea1a1096/attachment.pgp>


More information about the Koha-devel mailing list