-
Notifications
You must be signed in to change notification settings - Fork 55
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Support snapshot revert operation #3623
Conversation
bdc0316
to
61f9b17
Compare
61f9b17
to
017f330
Compare
Cockpit tests failed for commit 017f330. @martinpitt, @jelly, @mvollmer please check. |
Ech, can we please stop breaking rawhide in new an "interesting" ways every censored day? I'm still catching up with the fallout from the firewalld/selinux debacle..
Again, clearly not your fault. I'm still busy with investigating/reporting the other regressions from last night, so if someone else could have a look that'd be nice. However, as #3653 went green, it's worth retrying. |
The timing coincides with https://bodhi.fedoraproject.org/updates/FEDORA-2024-26a844bb1c . This also happens in podman PRs, e.g. containers/podman#23274 or containers/podman#23275 . Trivially reproducible in a rawhide VM with updating shadow-utils. With shadow-utils-2:4.15.1-8.fc41.x86_64 (not the latest version from that bodhi update, that's -9):
That most recent bodhi update reverted the change again:
and with that it works. So please retry. |
Cockpit tests failed for commit e5933ba. @martinpitt, @jelly, @mvollmer please check. |
I hacked around the shadow-utils b0rkage in cockpit-project/cockpit#20802 . Can you please retry the failed tests (or ignore the rawhide failures if you don't care)? |
e44d40e
to
e9c9985
Compare
rebased. This was just a conflict w/ the filesystem metadata PR, so it was not a lot of work to resolve the conflicts. |
03e19a3
to
12a2d19
Compare
Cockpit tests failed for commit 7d9664d. @martinpitt, @jelly, @mvollmer please check. |
Cockpit tests failed for commit 03e19a3. @martinpitt, @jelly, @mvollmer please check. |
Cockpit tests failed for commit 12a2d19. @martinpitt, @jelly, @mvollmer please check. |
f468a09
to
1d5489f
Compare
Another change is implemented. If reverting snap1_a into snap_1, the origin field is set to origin0, not to None as formerly. |
I think we should make the filesystem_destroy operation patch up the origin as well, to make it congruent with revert behavior. But that's more of a pain, because of the D-Bus. |
I think that makes sense. I'll try to get these changes tested in my snapshot manager VMs. |
21d5fd9
to
e4cba73
Compare
/packit build |
e4cba73
to
79c6bfd
Compare
Congratulations! One of the builds has completed. 🍾 You can install the built RPMs by following these steps:
Please note that the RPMs should be used only in a testing environment. |
1 similar comment
Congratulations! One of the builds has completed. 🍾 You can install the built RPMs by following these steps:
Please note that the RPMs should be used only in a testing environment. |
03f59f7
to
43fb5c5
Compare
Add MergeRequested property to filesystem D-Bus interface. Use it to set and unset the filesystem metadata field. Change destroy_filesystems to take into account the merge status of a filesystem. Move all the destroy functionality into thinpool implementation. Do more checks to avoid deleting filesystems that shouldn't be deleted. Check for situations where multiple snapshots are referring to the same deleted filesystem. Check for invalid scheduling requests when requesting or canceling a merge request. Add some tests.
Verify that scheduled merges are permissible. Return with an error without setting up any filesystems if duplicate UUIDs or names are found. Perform the merges before setting up any filesystems. If a merge can be rolled back, set up the two filesystems in the merge relation in the normal way. Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
If a snapshot was taken after the one reverted to, remove its origin field. If a snapshot was reverted, any snapshot that points at it should have its origin field removed. Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
Semantics preserving, so pass None where it is invoked. Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
Make this change visible on the D-Bus. Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
Signed-off-by: mulhern <amulhern@redhat.com>
To avoid a mutable snap and miscellaneous setting and unsetting in the setup method. Signed-off-by: mulhern <amulhern@redhat.com>
43fb5c5
to
1e3f6c0
Compare
Closes #3621
Closes #3622