There is pages in the admin on my site that currently don't load and there is a lot of users having issues loading a earning page. This site is heavily based on database so a expert in database is needed. This site was coded in a framework which is terrible. I need someone that can redo the site and handle hundreds of thousands of users. I have already spent a great deal of money on this other programmer helping keep the site running decent but he is expensive to redo this site. Here is some of the notes he suggested but I would like someone to redo most of it. Please contact me with any questions you may have and give me a rough quote and give me examples of sites you have done. Also any cookie-cutter pms will be ignored. The indexes I have added should make things quicker, and hopefully stop your crashes. The code making up your site is OK, but extremely bulky with lots of overhead, the use of a framework has obviously sped up development but most likely will continue to contribute to growing pains, I suggest a complete optimizing your server's MySQL config, a rewrite of the entire site, and replacing the SATA hard drive with a SAS 15k RPM drive. At the very least the "visitors" table should be broken up into several tables, which will require scrapping the current code logic completely. An UPDATE per page load is not good practice, you should use MySQL's INSERT DELAYED with a cron to update a cumulative stats table. For example, the site seems to track URL visit by IP, so you could have a visits table which is INSERT DELAYED into for each page load to store that person's IP, then use a cron every minute to update a "todaysIPs" table which tracks stats for IPs in the past 24 hours, each row would represent one IP with the IP being the index. Other columns could have the required time, date, and link URL data. Nightly you would summarize the "todaysIPs" data into your long term stats storage table which would contain all the viewing history data for each ip/URL, would grow in size daily, but wouldn't need to be queried thousands of times a day. This method would allow you to keep your "todaysIPs" data table short, allowing for fast searches / inserts when people view each link. I believe that if you stick with this code you will end up throwing a lot of money into hardware in order to grow.