Bug #63719
open[test] scribble()-based DiffIterate tests are too weak
0%
Description
scribble()-based DiffIterate tests are too weak: at least two regressions that should been caught by DiffIterate.DiffIterate or DiffIterate.DiffIterateStress were missed:
https://tracker.ceph.com/issues/50787
https://tracker.ceph.com/issues/63654
Aside from the randomness which can be both a good and a bad thing, asserts
https://github.com/ceph/ceph/blob/b27938f5f17898fc3da3d5af821957d6aee33501/src/test/librbd/test_librbd.cc#L7475
https://github.com/ceph/ceph/blob/b27938f5f17898fc3da3d5af821957d6aee33501/src/test/librbd/test_librbd.cc#L7648
https://github.com/ceph/ceph/blob/b27938f5f17898fc3da3d5af821957d6aee33501/src/test/librbd/test_librbd.cc#L7821
https://github.com/ceph/ceph/blob/b27938f5f17898fc3da3d5af821957d6aee33501/src/test/librbd/test_librbd.cc#L7902
ensure only that returned diff covers all changes that were made. If returned diff is too excessive or otherwise bogus, this isn't detected -- an implementation that always returns the entire range passed to rbd_diff_iterate2() as "new data" would pass all of these!
Updated by Ilya Dryomov 5 months ago
- Related to Bug #50787: rbd diff between two snapshots lists entire image content with 'whole-object' switch added
Updated by Ilya Dryomov 5 months ago
- Related to Bug #63654: [diff-iterate] ObjectListSnapsRequest's LIST_SNAPS_FLAG_WHOLE_OBJECT behavior is broken added
Updated by Ilya Dryomov 5 months ago
- Related to Bug #53897: diff-iterate can report holes when diffing against the beginning of time (fromsnapname == NULL) added
Updated by Ilya Dryomov 5 months ago
- Related to Bug #63771: discard at object end can grow the logical size of the object added