RAM usage on scalability tools using NTFS AND NTLM Authentication

[#1]

Hey just wondering we are using NTLM Auth and using NTFS Security, Is there anyway possible to test memory,  I can have 20 concurrent users on an app but the memory does not change .  however CPU Spikes high because it is intensive.  Thanks.

[Peter_Cammaert]

The memory test you want to perform, can it be destructive or non-destructive?

[#3]

Can you elaborate the difference please.

[Peter_Cammaert]

Memory tests are essentially one of two kinds:

1. Non-destructive tests: "what was there will stay there" - bytes are either only read or tested using a SAVE-WRITE-TEST-RESTORE cycle. RAM content is left untouched.
2. Destructive tests: "everything is lost" - bytes are tested by writing to/reading from without the test guaranteeing any particular RAM content. You better restart the machine afterwards.

Peter

[#5]

Well I am trying to create a JMeter test by having multiple concurrent users but currently my security is NTFS AND NTLM Authentication,  Since I can only have one user only run the test is there a way to have NTLM authentication and have multiple concurrent users to test if the server can handle multiple users?

[Peter_Cammaert · Accepted solution]

AFAIK that's called Load Testing or Stress Testing. Ignore my previous replies, I thought you were actually going to test memory.

I see your point now - and I don't think there is a simple solution without configuring N users instead of a single one, and assigning N whatever-CALs to those same users. The problem is that a single ID can have only one active session in the AccessPoint. The next user in your testing set with the same account will close/take over the session of the previous user with the same account. So the test is really limited to loading the first sheet and making some selections until the same user deactivates that session and starts a new one.

You could perform a manual test using a single user that tries to create two sessions and observe what happens with QVS allocated memory. If it does not increase, then all allocated memory (cache + document) is transferred to the second instance and 2 sessions or 100 won't really make a difference.(except for slowing down access). OTOH, if memory usage increases, then you can stress test server memory to a point with a single QV User+ CAL.

The community has a Qlik Scalability group, where Qlik people from the Scalability center may watch the discussions more often. Maybe you should post your question there as well.

Best,

Peter