Starting week number three at b5media and things are really starting to click.

I feel like I’ve settled in just a bit and I’m ready to get some real traction.  b5 is really growing and it’s great.  Over the last two weeks I’vehelped add a bunch of new sites to the network and we’ve got a number more coming in.

There’s a new theme that is being rolled out to a bunch of the blogs and I’ve helped with that, but due to the way it was implemented, we had some load problems as large common sections of html were being generated over and over again.  It was causing WAY more CPU usage than it needed to.

So, our excellent Linux guy Sean pointed me to a PECL module I’d not used before: memcached.

To quote http://en.wikipedia.org/wiki/Memcached:

memcached

 

From Wikipedia, the free encyclopedia

memcached is a general-purpose distributed memory caching system that was originally developed by Danga Interactive for LiveJournal, but is now used by many other sites. It is often used to speed up dynamic database-driven websites by caching data and objects in memory to reduce the amount the database needs to be read. Memcached is open source and released under a BSD license. It uses libevent.

Memcached lacks authentication and security features, meaning it should only be used on servers with a firewall set up appropriately. By default, memcached uses the port 11211.

Memcached’s APIs provide a giant hash table distributed across multiple machines. When the table is full, subsequent inserts cause older data to be purged in LRU order. Applications using memcached typically layer memcached requests and additions into code before falling back on a slower backing store, such as a database.

The system is used by several very large sites. Some well-known websites that use memcached are:^[1]

• LiveJournal
• Slashdot
• Wikipedia
• SourceForge
• GameFAQs
• Facebook
• b5media (Added)

It’s worked really well so far.  A few apache recompiles later and bam!  CPU load goes bye-bye.  With the initial partial rollout on a single channel we’ve tremendously reduced the load on the servers making 80 cache pulls per second instead of doing all of the db lookups and xml code generation.

I’m convinced!

I’m going to do some further implementation this morning we should be sitting sweet by EOD.

I’m also feeling good because I got the whole family up and did a morning workout in the new execise room I’ve setup in the garage.  I’ve also got Akaza hits nice and loud…

I’ve also got to send an email out to my Google Summer of Code student and we’ll get rolling on that this morning.  I’m looking forward to a good week.

Oh! and my pictures will be transfered by the end of the day so I can tell you all all about my toronto trip as I said I would two weeks ago…

Cheers all!