The online community for software testing & quality assurance professionals
Calendar   Today's Topics

Lost Password?


Software Testing >> Performance & Load Testing

Pages: 1

Reged: 11/07/12
Posts: 1
web apps performance testing :: LR vs LISA
      #719469 - 11/07/12 09:21 AM


we would like to know the positives and negative of using LISA for web apps based performance testing activity (this is not for SOA / web services, this is normal web apps via browser). our org wants to compare both LR and LISA and move forward as a standardized tool. we do have expertise in LR, but any insights or pointers on using LISA for a web application is appreciated.

a. does lisa provides the true client browser emulation?
b. how does it match to end-to-end transaction activity?
c. does lisa provides measurement on .js scripts execution metrics @ browser level


Post Extras: Print Post   Remind Me!   Notify Moderator  
James Pulley

Reged: 08/01/01
Posts: 5551
Loc: NC
Re: web apps performance testing :: LR vs LISA [Re: Gerind]
      #719473 - 11/07/12 09:55 AM

a) How does it matter to your application whether there is full web browser on the other end vs the request reproduction of what would be a full browser?
a.2) For LoadRunner you have several models if you need actual browsers: GUI Virtual User with your choice of any browser which runs under WIndows, TruClient variants which run on top of Firefox or IE layers, Citrix/RDP

b) Define end-to-end. There are probably a dozxen or so different definitions in the market depending upon which manufacturer is handing you the term, each claiming some truth in "true end-to-end"

c) Use a client side profiler for this in development and functional testing. The performance of this is too variable in real world to be considered a good measure. Add in a couple of toolbars, run it on an older browser, run in a crowded PC already polluted with a lot of background apps, run it in a RAM deprived PC space, all of which will result in different performance from what you would see in a test in a clean reference environment. As a part of solid performance engineering principals you should still measure it but functional and development profilers are going to be much more efficient in this space than trying to do this while n the middle of a performance test.

You may want to take a look at your application request model as well, for there are really very few in house true Asynch request applications which would demand full javascript execution in the browser for the in house application test. In contrast, there are a lot of third party apps (Google,etc...) where access to the third party component is absolutely asynchronous and would require either a reproduction of the Asynch code in your script or running Javascript in a full browser. But then you would not normally include the third party components in your performance test anyway - Can you imagine trying to coordinate with Google for a performance test of an application which includes Google Maps as an example?

Most applications which use Javascript are still reliably synchronous in their request behavior to the application under test across multiple iterations. I would recommend recording your application several times to check this behavior.

Lisa is a great tool for mocking third party interfaces, such as your credit card provider, etc... There is a place for both tools but I have not found that one is a direct replacement for the other.

Lisa was once for sale through HP VARs because of its strong interface reproduction capability. I would strongly encourage you to speak with both HP and ITKO directly and frankly about why HP decided to end the relationship with ITKO. I will leave HP and ITKO to go into the details of this and it likely will provide an additional data point for your product evaluation.

Post Extras: Print Post   Remind Me!   Notify Moderator  
Pages: 1

Extra information
0 registered and 8 anonymous users are browsing this forum.

Moderator:  AJ, RSBarber, rstens, Corey_G 

Print Topic

Forum Permissions
      You cannot start new topics
      You cannot reply to topics
      HTML is disabled
      UBBCode is enabled

Topic views: 2570

Rate this topic

Jump to

Contact Us | Privacy statement SQAForums

Powered by UBB.threads™ 6.5.5