search

Letractively / fuuka

Automatically exported from code.google.com/p/fuuka
Other
0 stars 0 forks source link

Implement memcached caching #63

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago 
An implementation of memcached would be nice, these search queries take some 
time and much longer if more than 1 person is using the archive.

Heres the CPAN page for it: 
http://search.cpan.org/~dormando/Cache-Memcached-1.29/lib/Cache/Memcached.pm

Original issue reported on code.google.com by aa...@installgentoo.net on 16 Sep 2011 at 11:40

GoogleCodeExporter commented 8 years ago 
They really don't, actually. Your issue is that the version you're using uses 
MySQL fulltext indexes for searching. This is really terrible and really 
inefficient. On easymodo, I'm using Sphinx (sphinxsearch.com), it performs 
amazingly well, I'm just finishing up some details before pushing it to SVN.

In addition, you can also make some MySQL tweaking. I recommend at least the 
following two tweaks on my.cnf:
low_priority_updates = 1
concurrent_insert=2

I'm not really sure what to cache with memcached even if I were to use it, 
though. MySQL's query cache does a good enough job of caching queries (as long 
as its cache doesn't get ruined by searches, as it is in your case). Memcached 
would only be useful if you had a dedicated MySQL server with a cluster of 
memcached servers or something. I doubt anyone is ever going to set up anything 
like that in order to archive 4chan.

I'll give you a bit to make your case for it, but otherwise, I'm leaning 
towards WontFix.

Original comment by eksopl on 17 Sep 2011 at 12:12

GoogleCodeExporter commented 8 years ago 
Suppose I don't have any.

Original comment by aa...@installgentoo.net on 20 Oct 2011 at 12:55

GoogleCodeExporter commented 8 years ago

Original comment by eksopl on 1 Nov 2011 at 5:18

GoogleCodeExporter commented 8 years ago

Original comment by eksopl on 19 Feb 2012 at 1:39