Wow okay, so is this like an HTTP based buffer pool manager? Instead of reading pages from disk it reads via HTTP?