tests: adapt disk assessment check #21739
Conversation
tomasmatus
force-pushed
the
fix-smart-test
branch
from
5ac9741 to
b2b4808
Compare
martinpitt
left a comment
martinpitt
left a comment
There was a problem hiding this comment.
Thanks! But I'm afraid this needs to do a libblockdev or m.image or TEST_SCENARIO or whatever dynamic check -- with this it's now failing on f42 in the opposite way.
tomasmatus
force-pushed
the
fix-smart-test
branch
from
b2b4808 to
e939502
Compare
test/verify/check-storage-smart
Outdated
| # latest libblockdev builds from copr have different behavior when assessing disk as failed | ||
| is_nightly_scenario = "daily" in getenv("TEST_SCENARIO", "") |
There was a problem hiding this comment.
This is not a robust condition. The API change can land in Fedora or updates-testing any day, or in a distro other than Fedora 41. This needs to be a feature check or a version comparison.
test/verify/check-storage-smart
Outdated
| if is_nightly_scenario: | ||
| check_smart_info("Disk is OK", "556 hours", "Aborted", bad_sectors="1") |
There was a problem hiding this comment.
I suggest to factorize this, do the version check and set some overall_status = "..." string conditionally there, then use the string here and below.
fixes: cockpit-project#21728 Libblockdev changed how SmartFailing property is set. Now it is less strict with when a disk is considered as failing. storaged-project/libblockdev#1097
tomasmatus
force-pushed
the
fix-smart-test
branch
from
e939502 to
7863c2c
Compare
martinpitt
left a comment
martinpitt
left a comment
There was a problem hiding this comment.
Thank you! How painful.. I would have liked the string factorization, but it's ok -- we'll see this again when we clean up the hack, and then we can do a simpler version check.
2 participants
fixes: #21728
Libblockdev changed how SmartFailing property is set. Now it is less strict with when a disk is considered as failing.
storaged-project/libblockdev#1097