Thresholds & Managing Application Performance Proactively

When ManageEngine came into being in 2003, OpManager was the first one to be released under its fold. The initial phases of Pre-Sales was exceptionally good and exciting. That was the beginning of my career as a Product Expert and Domain Consultant. We used to sit for hours with customers and give demos to meet their requirements. I know, that i am building stories here.  . Now, let me come to the point.

Customers ( or for that matter, everyone ) know(s) what Thresholds are. But, Re-arm value sounded different to them. I remember answering close to hundreds of support emails explaining what Re-arm values are. Re-arm value is that value of the attribute or monitor, which when reached, automatically clears the fault thus raised due to threshold violation.

Now, let us see how thresholds and re-arm values help in managing application performance proactively.

Threshold :

A baseline value for a parameter, which when monitored, indicates degradation of service when the monitored value is beyond the baseline value.

What does it indicate?

Threshold values are configured to ensure that systems are running fine and indicates optimum utilization of resources.Any violation in the Threshold value indicates a degradation in the performance of the application or server or network or all.

How do you configure thresholds to Monitored parameters in Applications Manager ?

For any monitored parameter, you can configure a Threshold value. You can configure Critical, Warning and Clear Threshold values for a parameter. To detail it, let us assume that you are monitoring CPU Utilization on a Web Server. You can configure to receive a Critical Event when CPU Utilization is greater than 90%, a Warning Event when CPU Utilization is between 60% and 90% and a Clear Event when CPU Utilization is less than 60%.

Threshold and Re-arm value
You can associate this threshold to all devices on which you monitor the CPU Utilization. Corrective Actions or Email / SMS Notifications can be configured at attribute level for Critical, Warning and Clear Alarms.

How to set meaningful thresholds?

Well, this is mostly dependent on your business requirements and environment and cannot be generalized. However, you can arrive a baseline value following the trend of these monitored parameters for 7 days or 30 days. Here is a 7 day report on CPU Utilization on a server, which help you arrive at a baseline threshold value.

7-day report on CPU Utilization in Applications Manager

Looking at the graph above, anyone can draw their own conclusion on setting a baseline value for CPU Utilization on the Server.

Suggested Reads for setting thresholds to MSSQL Monitor and Sun-solaris boxes :

1. MSSQL Monitor :

You can look at the document through the URL below for basic guidelines on setting thresholds for various parameters monitored on MSSQL.

2. Sun Solaris box :

See how to set thresholds through Sun System Manager and get them monitored through Applications Manager.

http://docs.sun.com/app/docs/doc/819-2666/6n4ujm6pv?l=sv&a=view

Well, dont ask me why this post came out as a Q&A, but this was supposed to be my first blog. Just as it happens in film releases of an debutante, Improving Network Discovery came out first ( may be, because my director loved the Network Discovery than this. Who knows, may be, he was not interested in this. ).

Kevin

Leave a Reply

Your email address will not be published. Required fields are marked *

*

HTML tags are not allowed.