Back before the iPhone, Droid and other smartphones came out with their fancy, full-featured web browsers, I decided to install a nice little WordPress plug-in called WordPress Mobile. The plug-in allowed my web site to be displayed in a friendlier way on mobile browsers. I was very pleased with this plug-in as it brought me visitors who would normally back out of my site, unable to view it in their dinky mobile browser. Little did I know this feature was costing me other visitors and money.
As my site became more popular, the web server was bogged down quite a bit. Sometimes it would go down for hours after running out of available RAM. DreamHost convinced me to purchase an upgraded PS account were you pay for the amount of RAM you use. I happily forked over more cash for the upgrade. Things improved, but I still had problems keeping my web server up. The RAM usage would spike at random times of the day and bring the server to a crawl. Visitors were backing out because load times were through the roof.
Frustrated, I finally started playing with the WordPress plug-ins to see if they were the culprit. I would disable them one by one and check to see if any server performance improvement was made. I found that when I disabled the WordPress Mobile plug-in, my RAM usage dropped by a quarter. I was shocked and felt dumb I had never tried this before. Here I was paying for 300MB at $30 a month. Without the plug-in installed, I just needed to pay $15 a month for 150MB of RAM usage.
It’s been a few days since I have disabled WordPress Mobile. I have lost a few mobile visitors, but my traffic is higher and the server never spikes or goes down anymore.
I’ve learned my lesson now. It doesn’t matter how cool a plug-in may be, I’ll need to check to see if it does more harm than good first. Some plug-ins aren’t designed well. Others are designed well but don’t mix with other plug-ins so well (which is what I think happened with WordPress Mobile). Next time I’ll do some testing before dragging that folder to my web server.