Configuring Server Manager (SM) monitor and alert mechanism-
Primary goal of any support engineer is to make sure that 99% of the times, when system breaks, he hears that from the server. And not from the user.
There are various tool sets that you can implement to achieve this..The one I will be talking about in this post is SM monitor/alert mechanism.
First step is to setup email SMTP server, Sender’s email ID in SM.
All JDE servers that need monitoring, need to be configured in SM (There are still some releases that allow not to have everything in SM!)
There are various parameters that can be monitored by SM agents. Some of the important one’s are –
a> Zombie Kernels
b> Outstanding Requests for kernels
c> Security Server error on HTML user login
d> Managed home/instance shutdown, restart
e> Long running batch jobs (have not been able to run this successfully yet..But it is important one)
Apart from the above, there are various environment specific alerts that one would like to setup; such as alerting based on specific error messages in the log files (on Enterprise server, JAS etc).
Unfortunately, log scraping is not yet available in SM (at least as of TR 184.108.40.206). Having this would help in getting alarm for scenarios like –
- When a user does something that consumes lot of memory on JAS, and generates out of memory error in the jas log.
- Call Object kernel goes zombie, and dependent error message shows up in the jdenet kernel.
- And so many more