Has Anyone Used Loadrunner with TM1

Post Reply
RickH
Posts: 1
Joined: Mon Apr 09, 2012 2:49 pm
OLAP Product: TM1
Version: 9.5.1
Excel Version: 2007

Has Anyone Used Loadrunner with TM1

Post by RickH »

We are looking at building a global budgeting application which may require some performance testing. We have used Loadrunner to do this with Hyperion and Essbase, but have no experice using loadrunner with TM1... I have been told that Loadrunner and TM1 will work together however IBM does not support it. Hoping there is someone out there that has used Loadrunner with TM1 and can give us some advise in this area.
Thanks
RickH.
nick_leeson
Posts: 98
Joined: Sat Feb 11, 2012 11:13 am
OLAP Product: TM1 9x, BPC, Hyperion, HANA
Version: TM1 10
Excel Version: Excel 2003 - 2010

Re: Has Anyone Used Loadrunner with TM1

Post by nick_leeson »

I am assuming you want to run LoadRunner against TM1 Web and I don't think that's a bad idea at all.

The point is what are you going to test

1) Concurrency of Users for TM1 Web.
2) Core TM1 will depend on rules, feeders and Cube Design. I don't think you will get much benefit of using LoadRunner for this.
3) User Concurrency for Locking etc. Again if you read some the posts you get a general idea for this.
4) Network Latency .

As for Excel Templates I have never used LoadRunner for Excel Sheets but this would mean opening the Excel Sheets and capturing data using LoadRunner (not sure how DBRW's will handle this ) but you can log how long the Excel Templates take to open etc etc.

At this years Forum in Melbourne there was I believe a query by a end user was raised for the need of a tool like Junit for Load Testing to IBM which I suspect will take a couple of years to materialize.


Cheers
gkatz
Posts: 4
Joined: Tue Feb 21, 2012 9:17 am
OLAP Product: Cognos Express
Version: 9.5
Excel Version: 2007 2010

Re: Has Anyone Used Loadrunner with TM1

Post by gkatz »

That's how we test cube performance for large amount of users:

1) fulfill total cube with appropriate dummy data
2) executing processes/reading/editing cube data via .NET API-created application for different number of dummy users. It is very easy to create C# multithreaded OOP application with different "action" classes for different dummy users, which can be simultaneousely executed for target cube.

So, stage 2 is just self-created Loadrunner analog. It is really simple and effective way for performance tests.
Post Reply