In phone networks, the difference between peak and average load is typically at least one order of magnitude higher. Every day there is a difference of 4x betw

In phone networks, the difference between peak and average load is typically at least one order of magnitude higher. Every day there is a difference of 4x between min and max, with the peak being sustained for multiple hours. That's why telco systems are sized as "BHCA" (busy hour call attempts) and "CAPS" (call attempts per second), which are peak rates, not average rates. Well, they are both averages, but not over the whole day or year! Then, as others have pointed out, there's a difference between a static site and a non-static one. This is mostly determined by I/O. A single disk can do between 100-150 IO operations per second (IOPS). Reading a row from a database will take several IOPS. In the systems I work with, we have to keep the entire database in memory to meet our performance numbers, and even then we are still O(database reads). Doing a SOAP request to a legacy platform that was designed to only support batch requests? Lots of opportunity for slowness. Particularly when you start to run into TCP timers controlling port reuse. Doesn't excuse the 404's though. That stuff is easy to test automatically.