Cascading test sets and optimal usage of host pool
Hi, we use Quality Center to perform regression tests.
We have about 1000 test cases which run every day in an automated manner. In order to make maximal use of parallelism (and to have them still organized somehow), we have created test sets of test sets, ie. a launcher which knows how to execute a test set. (Read: We launch the execution of a test set using a test case). This works quite well, except ...
The launcher has one drawback and I would like to discuss how to improve it. Basically every instance of the test set launcher reuses the same host pool ... And tries to grab the first shooter available in the list. Which somehow is always the same shooter and thus we get queued requests on one host, whereas the other hosts are idle. How can I optimize/force the usage of the other hosts?