- server hardware (memory, processors, hard drive arrays, etc.)
- network connections/bandwidth
- routers
- application software
Determining which of these factors is contributing to the degradation of overall performance can sometimes be difficult and the more data that is available, the easier it is make an accurate diagnosis. Unfortunately the traditional network diagnostics tools, such as, traceroute or ping may not help provide you with a complete picture because servers are more commonly behind firewalls. To learn more about traceroute check out the Wikipedia article on the subject.
However, there are solutions to the problem with new diagnostic tools, such as, Path Analyzer Pro which is traceroute on steroids. There is a free trail version you can download for Mac, Windows or Linux. It is easy to use, it graphically charts latency and you can export the data to a csv file. It uses a TCP NAT-friendly adaptive trace mode to perform a complete trace that the regular traceroute program could not via ICMP or UDP protocols.
Why is this beneficial? When you connect to a web server you are not making a direct connection between your computer and that computer. Instead there are numerous computers in between. For example, there are 14 hops or other servers or routers between my computer and the web server that this blog is hosted on. Tools like Path Analyzer Pro help you to see where the bottle necks are in the network. Is it your computer, the server on the over end or somewhere in between. You can do a single trace, timed trace or continuous trace with Path Pro Analyzer and it graphically displays via a number of charts the minimum, maximum and average latency or delay at each point along the way.
Here is a screenshot of my connection to the Alumnae Association web server:
