Repop and lost-unfound overhaul #7765
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
qa-suite PR: ceph/ceph-qa-suite#832 Test run: samuelj-2016-02-22_06:57:34-rados-wip-sam-working-distro-basic-smithi Ready for merge. |
| RepGather *simple_repop_create(ObjectContextRef obc); | ||
| void simple_repop_submit(RepGather *repop); | ||
| OpContextUPtr simple_opc_create(ObjectContextRef obc); | ||
| void simple_opc_submit(OpContextUPtr ctx); |
There was a problem hiding this comment.
all of the other methods use 'ctx' to mean OpContext (e.g., finish_ctx).. any specific reason to use opc here?
There was a problem hiding this comment.
These two are sort of external interfaces to the other ctx methods. I could switch it if you feel strongly.
athanatos
changed the title
| @@ -1415,6 +1417,23 @@ class ObjectStore { | |||
| } | |||
| data.ops++; | |||
| } | |||
| void try_rename(coll_t cid, const ghobject_t& oldoid, | |||
| const ghobject_t& oid) { | |||
| if (use_tbl) { | |||
There was a problem hiding this comment.
do we already have a barrier feature prevent "use_tbl" is true?
athanatos
force-pushed
the
wip-lost-unfound-repop
branch
2 times, most recently
from
c6c9876
to
b3c6117
Compare
Signed-off-by: Samuel Just <sjust@redhat.com>
This way, we don't expose the RepGather structure up into the users (which pretty much exclusively use the repop->ctx member anyway). This will pave the way to removing RepGather::ctx. Part of this involves generalizing the repop callback members (queue_snap_trimmer and on_applied) to on_applied, on_committed, on_success, and on_finish on both OpContext and RepGather. Signed-off-by: Samuel Just <sjust@redhat.com>
execute_ctx is the only path which actually has a client op, let's clarify eval_repop by moving that logic inline there. This also eliminates most of the repop->ctx users from eval_repop. Signed-off-by: Samuel Just <sjust@redhat.com>
We'll need this shortly when we remove do_osd_op_effects from eval_repop. Signed-off-by: Samuel Just <sjust@redhat.com>
This is better anyway, do_osd_op_effects on a non-client op was somewhat silly without an op (even if you filled in the other fields like watch_connects, it would just silently skip it). It also eliminates another dependence of RepGather on OpContext. Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
…evict Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
…saction submit_transaction takes ownership of the transaction implicitely. Make this implicit using an rvalue ref to force usage of std::move and move constructor. Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
…opcontext This gets us one step closer to removing OpContext from Repop. Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
We want these to be processed ahead of new client ops since there are resources being held. Fixes: 14313 Backport: hammer Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
This way, we can get proper logging within static functions. Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
This should not change behavior at all. It'll be useful in the next few commits. Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
athanatos
force-pushed
the
wip-lost-unfound-repop
branch
2 times, most recently
from
502fbe5
to
8938eaa
Compare
Similar to collection_move_rename, except we ignore ENOENT and don't allow different collections. Signed-off-by: Samuel Just <sjust@redhat.com>
Signed-off-by: Samuel Just <sjust@redhat.com>
…e safely Using a MOSDPGLog was unsafe since it is not ordered with respect to repops. Instead, use a new message sent through the same paths as repops. Signed-off-by: Samuel Just <sjust@redhat.com>
athanatos
force-pushed
the
wip-lost-unfound-repop
branch
from
8938eaa
to
e7edf20
Compare
athanatos
changed the title
|
Passed tests after rebase: @liewegas Ok to merge? |
|
yep! On Mon, 29 Feb 2016, Samuel Just wrote:
|
| pinfo.last_update = info.last_update; | ||
| pinfo.stats.stats_invalid = true; | ||
| } | ||
| for (auto &&i: entries) { |
There was a problem hiding this comment.
&& does not make i a rvalue reference.
athanatos
added a commit
that referenced
this pull request
Feb 29, 2016
Repop and lost-unfound overhaul Reviewed-by: Sage Weil <sage@redhat.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
mark_lost_revert/delete can race dangerously with normal IO due to not being ordered with other repops. This series generalizes the repop ordering machinery and introduces a new message for performing mark_unfound_revert/delete.