Some questions after testing AppDynamics Lite
Dear AppDynamics community,
I have now played around a little bit with the Appdynamics Lite product.
We use for our software an inhouse build application server based on the Jetty webserver.
Wenn trying to do some performance tests I had some problems to get good figures for all scenarios.
On our server we have different types of requests. For example user requests from the Client application.
These are many small requests going from the client via the webserver to the business layer and the database.
On the other hand we have some webservices running on the application server. Very few requests but running very long as they produce XML files on request.
With this scenario I tried to monitor the system performance. My problem is now that I find a good way to monitor the long running webservice requests because AppDynamics seems to need an minimum of n requests per minute but some of them are running longer as a minute and I found no good way to inspect them.
The other problem is that the threashold values seem to be global and not per business transaction. But for one transaction 2 seconds is slow (client requests) and for others 2 seconds is nothing (webservices).
Is there a way to control this per business transaction? Or is there at least a way to ignore the thresholds for some of the known long running transactions?
For comparing performance is there a good way to force AppDynamics to capture the next executed business transaction? This would be very usefull to test different performance settings for the webservices.
Executing them a few dozend times to get some captured snapshots is time consuming as they run very long.
And finally some question regarding the pro version. Is it possible to use this to do performance monitoring on a customer site. For example temporary install the software on a customers site. Do some performance monitoring with their data and then uninstall it again when we have identified the bottleneck. Is this allowed license wise or is a pro version always bound to a specific installation of an application server and every customer is forced to buy his own version?
Thanks for your replies,
- Bernd Rosstauscher
I have now played around a little bit with the Appdynamics Lite product.
We use for our software an inhouse build application server based on the Jetty webserver.
Wenn trying to do some performance tests I had some problems to get good figures for all scenarios.
On our server we have different types of requests. For example user requests from the Client application.
These are many small requests going from the client via the webserver to the business layer and the database.
On the other hand we have some webservices running on the application server. Very few requests but running very long as they produce XML files on request.
With this scenario I tried to monitor the system performance. My problem is now that I find a good way to monitor the long running webservice requests because AppDynamics seems to need an minimum of n requests per minute but some of them are running longer as a minute and I found no good way to inspect them.
The other problem is that the threashold values seem to be global and not per business transaction. But for one transaction 2 seconds is slow (client requests) and for others 2 seconds is nothing (webservices).
Is there a way to control this per business transaction? Or is there at least a way to ignore the thresholds for some of the known long running transactions?
For comparing performance is there a good way to force AppDynamics to capture the next executed business transaction? This would be very usefull to test different performance settings for the webservices.
Executing them a few dozend times to get some captured snapshots is time consuming as they run very long.
And finally some question regarding the pro version. Is it possible to use this to do performance monitoring on a customer site. For example temporary install the software on a customers site. Do some performance monitoring with their data and then uninstall it again when we have identified the bottleneck. Is this allowed license wise or is a pro version always bound to a specific installation of an application server and every customer is forced to buy his own version?
Thanks for your replies,
- Bernd Rosstauscher
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Categories
- All Categories 235
- Ideas 19
- Problems 139
- Questions 71
- Announcements 1
- Useful Tips 0
- FAQ 5
Comments
Kind regards, -Dee
You can also try our pro version of product which has concept of Diag session which can BT Level thresholds .
Regards,
Arpit Patel