topic Re: Qlik Sense Scalability tool purpose question. in Integration, Extension & APIs

Qlik Sense Scalability tool purpose question.

M_Bilto — Sat, 03 Nov 2018 12:18:50 GMT

Hello Qlik community

Recently in one of our projects we were asked to test the capabilities of an environment to handle about 1000 concurrent users. so my boss found the Qlik Sense scalability tool and told me to try to use it so we can stress test the environment.

Though I believe he had the wrong Idea about the tool? The way he thought this tool works was that you can, before you purchase any licenses for user tokens, test a thousand number of users so that the clients can decide how much they should buy. The way I think it works is that if you have enough tokens, you can test how well they fare in your current environment.

my question is, do we need to have enough users in the first place in order to create a test with this tool? or can it generate virtual users for the test purpose that would tell us "Hey look, this is how well your environment runs with 1000 users at the same" before we get 1000 tokens?

thank you in advance for clarification and help.

Re: Qlik Sense Scalability tool purpose question.

Levi_Turner — Mon, 05 Nov 2018 19:25:19 GMT

Hey Mazen,

It can do both (generate random users, emulate specified users). Out of the box, the prefix method will prefix the user creation with the prefix specified:

Hope that helps

Re: Qlik Sense Scalability tool purpose question.

M_Bilto — Sun, 02 Dec 2018 09:10:35 GMT

Hello Good Sir,

thanks for your reply i appreciate your time. Indeed thats what it did, however my question was, say i want to use a stress test with 50 users. on the qlik sense server, I need to have enough tokens to pull that off right?

if i want to use 50 concurrent users i need to have 50+ tokens when i enter the QMC license management correct?

thanks in advance.

Re: Qlik Sense Scalability tool purpose question.

Jens_Argentzell — Mon, 03 Dec 2018 08:14:47 GMT

Hi,

Yes, with the Scalability tool you simulate virtual users to understand how the system coupe with different loads and usage patterns. For the environment you run simulation towards it is no difference it is a real user or virtual simulated user. Both consume the same amount of tokens.