As test implementation reveals in [1], these are the following steps:
# create an instance from volume
# write content to volume on instance
# delete instance <<==
# create a 2nd instance from volume
# check the content of written file
# snapshot a volume
# create a 3rd instance from snapshot
# check the content of written file
The step outlined above seems to be the one that causes the failure (test fails on [2]). This would indicate that once the namespace is cleared (thanks to change 112465), it is no longer re-configured correctly.
More triaging:
As test implementation reveals in [1], these are the following steps:
# create an instance from volume
# write content to volume on instance
# delete instance <<==
# create a 2nd instance from volume
# check the content of written file
# snapshot a volume
# create a 3rd instance from snapshot
# check the content of written file
The step outlined above seems to be the one that causes the failure (test fails on [2]). This would indicate that once the namespace is cleared (thanks to change 112465), it is no longer re-configured correctly.
[1] - https:/ /github. com/openstack/ tempest/ blob/master/ tempest/ scenario/ test_volume_ boot_pattern. py#L155 /github. com/openstack/ tempest/ blob/master/ tempest/ scenario/ test_volume_ boot_pattern. py#L163
[2] - https:/