Closed trenxiaof closed 5 months ago
The expected results:
disk_pool is None: bfv + swap + not swap_force_mdisk --- -->create by vdisk bfv + swap + swap_force_mdiks --->fail dasd + swap ->fail disk_pool is not None: bfv + swap + swap_default_with_mdisk --> create by mdisk bfv + swap + not swap_default_with_mdisk -->create by vdisk dasd + swap + swap_default_with_mdisk ---> create by mdisk dasd + swap + not swap_default_with_mdisk -->create by vdisk
@jichenjc @SeanHQF Did above tests, except for swap only case(it hungs for the issue:https://github.ibm.com/zvc/planning/issues/17425), all the other cases are as expected(keep the current default behavior and can controled by the new config swap_default_with_mdisk)
will continue to test swap only case.
@trenxiaof this is great, can you help fix the failed test then we can merge this
please hold on the merge!!
test the swap only case on non-ssi host, all the results are OK On ssi host, sometimes the deployement is OK, sometimes failed for the placement update problem should be the same reason as https://github.ibm.com/zvc/planning/issues
so we can merge the pr first. Thanks @jichenjc @SeanHQF
https://github.ibm.com/zvc/planning/issues/10607
previous config: swap_force_mdisk: False only works for swap only case( disk_pool is None, bfv + swap)
new config: swap_default_with_mdisk: True applies for all the cases when disk_pool is not None (including bfv+swap), if True, create swap with mdisk, if False, create swap device with vdisk