Qlik Sense Enterprise Scalability Tools

[Daniel_Larsson]

Note: Our legacy performance testing offering can be found here: Qlik Sense Scalability Tools.

The Qlik Sense Enterprise Scalability Tools are used to create and run load and performance tests towards Qlik Sense Enterprise for Windows (QSEfW).

Included parts are:

• Documentation on how to use the package
• Command line tool for Windows, Mac and Linux. This command line tool (gopherciser) is open source and can be found here .
• GUI for Windows, Mac and Linux.

Downloadable files can be found here:
📎 Qlik Sense Enterprise Scalability Tools 

Qlik Sense documents to help analyze result and log files can be found here :https://community.qlik.com/docs/DOC-15451

Change log

v1.12.0

• Updated scheduler
• Gopherciser v0.18.5 - v0.18.7
  • performance enhancements
  • new action stepdimension

v1.11.0 - v1.11.1

• Gopherciser v0.17.26 - v0.18.5
  • Bugfixes

v1.10.2 - v1.11.0

• Support for Feb 2024 Sense release
• Gopherciser v0.17.21 -> v0.17.26
  • Bugfixes
  • Report sub hypercube errors as errors instead of warnings (unless calculation condition)

v1.10.1 - v1.10.2

• Rebranded with new Qlik Logo

v1.10.0 - v1.10.1

• Gopherciser v0.17.7 -> v0.17.21
  • Added support for layout container object
  • Added support for new pivot table
  • Added support for sn-text object
  • Added support for sn-table object
  • Added support for sn-shape object
  • Added hyper cube sub-errors as warnings (These warnings will become errors in v1.11.0)

v1.9.0 - v1.10.0

• Gopherciser v0.17.7 -> v0.17.11
  • Open Hub Api compliance August 2023 release
  • Open App Api compliance August 2023 release

v1.8.0 - v.1.9.0

• Updated scheduler
• Updated reglog
• Gopherciser v0.17.4 -> v0.17.7
  • Changed parameters for prometheus push gateway
  • bugfixes

v1.7.0-v1.8.0

• Updated scheduler
• Gopherciser v0.17.0 -> v0.17.4
  • Configurable max websocket frame size
  • Improved disonnect discoverability
  • Add in-mem Get app structure method
  • [Apicompliance] openapp
  • Bugfixes

v1.6.0 - v1.7.0

• New installer - Option for personal install which should solve issues with "Administrator" privileges.
• Updated scheduler
• Gopherciser v0.16.2 -> v0.17.0
  • Bugfixes
  • Add option to print result json to script validate command
  • Updated api compliance for openhub action.

•  

•  

Your use of Qlik Sense Enterprise Scalability Tools will be subject to the same license agreement between you and Qlik for your Qlik Sense License. Qlik does not provide maintenance and support services for the Qlik Sense Enterprise Scalability Tools, however please check QlikCommunity for additional information on use of these products.

[sri_c003]

Thank you. Since I was only doing parallel user testing, I set the iterations from 25 and 1 leaving the concurrent users at 25. This time I didn't get any max sessions used error since every user is used only once.
I am still getting the "ERR failed to get DocList: EOF Action=openhub"; not a showstopper though.

[Daniel_Larsson]

---

@sri_c003 wrote:
I am still getting the "ERR failed to get DocList: EOF Action=openhub"; not a showstopper though.

---

This error is most likely due to authentication problems, your settings for the server etc is correct and a websocket is established, but on first command we are instantly disconnected without error.  Assuming you are using header authentication? In that case make sure the headername corresponds to the header name defined in your virtual proxy.

Also please also please verify you can log in "manually" as that user towards the virtualproxy using a modify header extension in your browser. E.g. if using chrome, https://chrome.google.com/webstore/detail/modheader/idgpnmonknjnojddfkpgkljpfnnfcklj . There are many others, same for firefox.

[sri_c003]

Excuse me for me ignorance here, but none of the users I used show up under QMC - I felt these were virtual users with no real existence. Do all these users need to exist on Qlik and need to show up in QMC?

To add, none of these users have showed up in the tool command run output as unable to login or anything like that. I found the logs from the time when i run the tool and all entries show up as the root user --> DEV1623\qlikadmin.

[Daniel_Larsson]

---

@sri_c003 wrote:

Excuse me for me ignorance here, but none of the users I used show up under QMC - I felt these were virtual users with no real existence. Do all these users need to exist on Qlik and need to show up in QMC?

To add, none of these users have showed up in the tool command run output as unable to login or anything like that.

---

If you can run a test without errors yes they should show up, as far as Sense knows they are real users. But as I pointed out in previous post you seem to have an authentication issue and the users are not actually let in.

[sri_c003]

Thanks @Daniel_Larsson. I shall run the test again and check.

[sri_c003]

I am able to run the tool and see the results as well as the results log. But the traffic logs are not being generated.

Help please.

[sri_c003]

Thanks @Daniel_Larsson. I got the tool working.

I am able to get the tool log file; but in spite of me telling the tool to generate traffic logs, it is not generating any tsv file.

[SebastianFabian]

Hi! The tool does not generate a separate traffic log - the traffic logging is added in the regular result log. So take a look in there and you should find additional log lines marked with traffic in addition to the regular logging.

[Daniel_Larsson]

In your screenshot you have clearly named the file *.log so in your case you would have a file called .log in the path you defined. This is a  .tsv file, if you want it to work in the analyzer you need to save it with .tsv extension.

[Daniel_Larsson]

And just to note, We normally do not recommend having full traffic turned on if you want to analyze performance as this would create a very big result file, and you would most likely performance test your harddrive instead. For performance analysis using the analyzer no traffic is needed. If you want for some reason to see the response time of individual requests you can turn on the traffic metrics instead of traffic. The only reason to turn on traffic would be to see all sent and received requests including everything in the result file, i.e. mostly for debugging.