Skip to content
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

rbd-mirror: primary image should register in remote, non-primary image's journal #18136

Merged
merged 2 commits into from Oct 10, 2017

Conversation

dillaman
Copy link

@dillaman dillaman commented Oct 5, 2017

No description provided.

@dillaman dillaman changed the title rbd-mirror: primary image should register in remote, non-primary image's journal [DNM] rbd-mirror: primary image should register in remote, non-primary image's journal Oct 5, 2017
@dillaman dillaman changed the title [DNM] rbd-mirror: primary image should register in remote, non-primary image's journal rbd-mirror: primary image should register in remote, non-primary image's journal Oct 6, 2017
Copy link
Contributor

@trociny trociny left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

lgtm

@trociny
Copy link
Contributor

trociny commented Oct 8, 2017

@ceph-jenkins retest this please

} else if (m_local_image_tag_owner == librbd::Journal<>::LOCAL_MIRROR_UUID) {
dout(5) << "local image is primary" << dendl;
on_start_fail(0, "local image is primary");
return;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@dillaman Running stress test locally I observed a false failure, when an ImageReplayer failed on start with the status "state=up+error, description=remote image does not exist", while it was expected to be "state=up+stopped, description=local image is primary".

The sequence was following:

  1. a mirrored image (primary) is created in cluster2;
  2. the local pool watcher for cluster2 mirror daemon gets notification about the new image
  3. cluster2 mirror daemon starts the image replayer
  4. start fails in prepare_remote_image because the image does not exist in the remote cluster yet.

If my explanation is not clear or you would like to see details I can provide logs.

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@trociny Thanks -- good catch. Update pushed to ignore the errors from prepare remote if the local image is primary.

Jason Dillaman added 2 commits October 9, 2017 20:53
The client will now be registered before the bootstrap state machine
is invoked.

Fixes: http://tracker.ceph.com/issues/21561
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
@trociny trociny merged commit 069c8dc into ceph:master Oct 10, 2017
@dillaman dillaman deleted the wip-21561 branch October 10, 2017 15:07
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
2 participants