I’ve been writing blogs for nojitter.com for the past several months. I thought that links from here would be useful so that anyone following my blog posts can find them too.
The blog post that I did in December was about detecting network packet loss. It was based on working with a couple of customers where there was significant application slowness. In one case it was a video conferencing system and in another case it was business applications.
It was easy to determine that there was packet loss in the video case. The video was crummy. The task then turned to determining why. Read the article (see link below) to see what caused it.
In the second case, we were doing a network assessment because there were significant complaints that applications were slow. The design part of the assessment turned up a few things, but nothing really significant. However, the network performance part of the assessment found that some key infrastructure links were dropping significant numbers of packets as well as a lot of edge ports with duplex mismatch. While working on this case, I decided that I need to start looking at raw interface error/drop counts as well as percentages. Read the article to see why.
Both stories are in the article Detecting Network Packet Loss.