CODEWITHSUNDEEP
monitoringloadrunner
Magnus Jensen
Magnus Jensen

Reputation: 993

sitescope monitors to include when running loadrunner test

What are the common monitors to include in a sitescope set-up when running loadtests with LoadRunner. I have asked for CPU, Memory and Disk I/O for my Linux instances running JBoss in a cloud based environment. Is it possible to add monitors for i.e. database connection pooling, to monitor number of connections to the database during loadtesting? What about monitoring other metrics? Anyone having set up sitescope for the purpose of monitoring during loadtesting with LoadRunner+ And if yes what did you decide to monitor?

Upvotes: 1

Views: 1611

Answers (1)

James Pulley
James Pulley

Reputation: 5692

The monitors are clearly related to your application logical and physical architecture. The OS level items are likely to be common across all hosts, but then when you get to the app server tier, such as a Java App Server or a database, you would potentially be pulling deep diag for your JVM or the V$ tabkle set values (in the case of ORACLE) or system table information from MSFT/SYBASE SQL Server.

I look at it this way philosophically. You have resources which are finite on every host and in every service. Monitor the use of these finite resources. Next, you have health level stats which are layered above the basic finite resource set. Collect the top 10 of these on each service. Reserve the deep probing metrics for when you need to examine in a particular area.

So, yes, you can monitor connection pooling to the database. Depending on what path you want, you can either look at it from the JMX side of the house or your can look at it from the database side of the house.

Upvotes: 1

Related Questions