Some intermittent failures might be caused by running our tests on uderpowered clusters.

We should find a way how to have our openstack flexy-install provisioned clusters have more resources available.

We have seen complaints about small disk-space and some pods being evicted after end of the tests

• Try figuring out how to bump the flavour of undreslying vms to have bigger memory and cpu count with  z-stream test pipeline and note the test stability.
• Try bumping number of worker/master nodes for a cluster with z-stream pipeline and note the test stability

If stability improves, consider updateing the defaults.