It's all about the answers!

Ask a question

RQM: How to use Multiple Command Line Adapters on a single Machine with different/limited resources


Stefan Friedrich (133) | asked Jul 04 '16, 12:18 p.m.

Is it possible to model licenses/resources per CmdLineAdpter on a single Machine, where each of the 
CmdLineAdapter starts the same program using different arguments (and hence uses different licenses).

Currently, I managed to perform a automatic selection based on the IP address (Machine) only,
but this would require duplicating Machines/VMs.

Any ideas ?

Thanks, Stefan

Comments
Don Yang commented Jul 05 '16, 12:54 a.m. | edited Jul 05 '16, 12:54 a.m.

Hi, Stefan

It is not clear what you mean by "each of the 

CmdLineAdapter starts the same program using different arguments", please provide some examples on how you defined the CLI adapters.
In my understanding, if there are multiples CLI adapters on one machine, I am able to run all of them(different name and id) and allocate them to different test script/test case and run them at the same time. I am not sure what you by "require duplicating Machines/VMs", please clarify.


Stefan Friedrich commented Jul 05 '16, 2:15 a.m.
HI Don,

Each automated/remote test script starts a local python interpreter via cmd.exe using the CLI adapter as defined in the section "Use test resources that are local to a test machine": e.g.


  command = cmd 
  arguments = /c " c:\bin\rqm-start-testTE.bat <testscript_nr>.py "

When I start multiple CLI adaptes (with different names & IDs) on one machine, they are associated with the IP/name of that one machine. 
I am able to edit the resources of that machine, but I cannot edit the resources of the CLI adapters which would be necessary for the selection according to the TestEnvironment (resource) definition.
Hence, I added a few machines for concurrent test executions (which are basically duplicates of that one machine but have different IPs).


One answer



permanent link
Thomas Rauscher (1111) | answered Jul 08 '16, 11:17 a.m.
We try to accomplish that one test adapter is responsible for tests requiring a dongle-locked library, so these tests must be started at most once.

All other tests can run freely without any licensing limitations in parallel and could use the second test adapter.

I think the problem is that the resources (has dongle or not) are mapped to the machine (or its IP address), not to a particular test adapter instance. So we cannot model two test adapters with different available resources, as the resources are bound to the machine entry.

So the only current way would be to use additional physical machines / VMs to get additional instances of the test adapters on different machines/IP addresses. But that seems heavy-weight for the purpose.

Or is there something we completely miss ...?


Your answer


Register or to post your answer.


Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.