RIGHT Blogs                                                              

Application Health Check Dashboard and Alerts

Having been part of new development and applications under support and maintenance for various applications for so many different companies, a question that is frequently asked -  how do we solve issues proactively?  The answer seems to have a lot to do with "Application Health Check Dashboard and Alerts".

Our recommendation to our clients is to implement this within their corporate infrastructure as best practices per application. What are the needs of such a system ? The criteria is really simple and does not have to be fully defined from day zero. Here are some of the requirements that will get us there.

1. Define application specific metrics that can be tracked on a daily/weekly/monthly/quarterly/yearly basis.

2. Identify what are acceptable thresholds for each of these metrics. At the minimum identify what is consider green/yellow/red threshold ranges.

3.  Design and implement a interactive web UI dashboard which will host these application specific metrics where they can be quickly checked on a on-demand basis. Ideally this should be accomplished by another corporate application which keeps track of these statistics and displays them in a standard format.

4. Identify the alerts mechanism where multiple levels of (mailing list) groups can be notified if the thresholds are breached. Again this is ripe to be part of a corporate application which would keep track of these notifications on a active basis.

This corporate application which would maintain the Application Health Check data, dashboard and alert management can be implemented with so many off the shelf tools like Nagios, and ZenOSS to name a few. The important part is to have each application report their statistics to this centralized repository in a standardized format.

Have you done this differently ? Please feel free to discuss.

Regards, Sandeep