Re: Re: algorithm help for determining efficient db data retrieval

[AidanLee]

The Lazy Population section of the Class::DBI docs was enlightening. Ideally i'd like to be able to hang onto the top x most frequently used objects in memory, but at the moment I live in a cgi world. That will be a possibility when I migrate the platform to mod_perl, but for now I'm stuck with limited-to-no caching between requests.

[tachyon]

Not true. Your RDBMS will cache all the most recently accessed data for you in memory without you lifting a finger. If you feed it more memory it will cache more. Here is a sample my.cnf that we use on servers with 2GB of RAM running squid and apache as well. Compared to the default settings throughput is about tripled for the way we use it.

[root@devel3 root]# cat /etc/my.cnf
[client]
socket=/tmp/mysql.sock

[mysqld]
datadir=/var/lib/mysql
socket=/tmp/mysql.sock
#set-variable=wait_timeout=3600
set-variable=key_buffer=640M
set-variable=max_allowed_packet=32M
set-variable=table_cache=512
set-variable=sort_buffer_size=32M
set-variable=record_buffer=32M
set-variable=read_buffer_size=32M
set-variable=myisam_sort_buffer_size=64M
set-variable=thread_cache=8
set-variable=query_cache_size=32M
set-variable=tmp_table_size=32M

[snip]