Jesse Becker has written an excellent article this week for Linux Magazine on how to successfully monitor a high performance computing cluster. The ‘HPC’ adjective is very important here. There are plenty of open source and commercial monitoring packages available to the consumer. Most of which, however, are based solely on enterprise requirements. The HPC scene often requires different techniques, metrics and time goals.
Jesse is a sysadmin in Maryland, so he is working with the tools on a daily basis. The article is full of great examples and lots of quality meat.
I won’t steal much more of Jesse’s thunder, so head over to Linux Magazine and read the full article here.