09-25-2013 05:58 AM
I'm a CM newbie and I'm trying to understand what are some common metrics I should be watching to monitor applications and even platform health in CM. I'm in a Prod Support role, but I'm not an SA on any of the servers. I'm basically trying to determine which charts do people use on a regular basis to monitor the health and performance of applications running on the cluster.
10-02-2013 06:39 PM
The CM UI will actually provide visual hints on what needs attention. Most of the time the graphing is relvant as you want to observe performance during job activities in map reduce, etc. For example within a service role (like HDFS jobtracker, or the NameNode) Disk performance of specific nodes can be a good indicator. If you are seeing a high level of wait times, that the node could be overloaded, for example.
We have a free web based training on CM available here, it will help you have a better understanding of CM and its use: