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

objecter: can get stuck in redirect loop if osdmap epoch == last_force_op_resend #4597

Merged
1 commit merged into from Jul 21, 2015

Conversation

Projects
None yet
3 participants
@xinxinsh
Member

xinxinsh commented May 7, 2015

@xinxinsh xinxinsh added this to the firefly milestone May 7, 2015

@xinxinsh xinxinsh self-assigned this May 7, 2015

@ghost

This comment has been minimized.

ghost commented May 22, 2015

@xinxinsh how come the author of aced7aa is you although the signed-off-by are Jianpeng Ma and Sage Weil ?

@xinxinsh

This comment has been minimized.

Member

xinxinsh commented May 24, 2015

@dachary ,jianpeng directly modify on my branch, is there any problem?

@ghost

This comment has been minimized.

ghost commented May 24, 2015

@xinxinsh it's not a problem, I was just curious if it was an accident or intended.

xinxin shu
osdc: add epoch_t last_force_resend in Op/LingerOp.
Using this field record the pg_poo_t::last_force_op_resend to avoid op
endless when osd reply with redirect.

Fixes: #11026
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit def4fc4)
@smithfarm

This comment has been minimized.

Contributor

smithfarm commented Jun 8, 2015

@xinxinsh: could you add

(cherry picked from def4fc4ae51174ae92ac1fb606427f4f6f00743e)

to the commit? This makes it identifiable as a backport.

@xinxinsh

This comment has been minimized.

Member

xinxinsh commented Jun 8, 2015

@smithfarm , updated

@ghost

This comment has been minimized.

ghost commented Jul 8, 2015

@smithfarm @xinxinsh it conflicts with #5043, could you please make it so they do not ? I will help with integration :-)

@ghost ghost changed the title from objecter: can get stuck in redirect loop if osdmap epoch == last_force_op_resend to [DNM] objecter: can get stuck in redirect loop if osdmap epoch == last_force_op_resend Jul 8, 2015

@xinxinsh

This comment has been minimized.

Member

xinxinsh commented Jul 9, 2015

@dachary , i will take a look

@xinxinsh

This comment has been minimized.

Member

xinxinsh commented Jul 9, 2015

@dachary , @smithfarm , merge conflict is introduced by variable last_force_resend in PR 4597 and variable ctx_budgeted in PR 5043, these two variables are added in the same location, can we just add new commit to resolve this conflict when we merge these two PR, or adjust the location of one of variable ?

@ghost

This comment has been minimized.

ghost commented Jul 9, 2015

@xinxinsh @smithfarm I think the right way to deal with this is to wait for one of them to be merged and then deal with the others. I'll send a mail to the list, it's worth a discussion.

@ghost ghost changed the title from [DNM] objecter: can get stuck in redirect loop if osdmap epoch == last_force_op_resend to objecter: can get stuck in redirect loop if osdmap epoch == last_force_op_resend Jul 9, 2015

@ghost

This comment has been minimized.

ghost commented Jul 21, 2015

the objecter is used everywhere, http://pulpito.ceph.com/loic-2015-07-09_21:09:04-rados-firefly-backports---basic-multi is green, meaning this backport is ok to merge and see more tests.

ghost pushed a commit that referenced this pull request Jul 21, 2015

Loic Dachary
Merge pull request #4597 from xinxinsh/wip-11026-firefly
objecter: can get stuck in redirect loop if osdmap epoch == last_force_op_resend

Reviewed-by: Loic Dachary <ldachary@redhat.com>

@ghost ghost merged commit 8f11ab7 into ceph:firefly Jul 21, 2015

This issue was closed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment