Merge lp:~abentley/launchpad/same-queue-query-fmr into lp:launchpad
| Status: | Merged |
|---|---|
| Approved by: | Aaron Bentley on 2012-08-07 |
| Approved revision: | no longer in the source branch. |
| Merged at revision: | 15760 |
| Proposed branch: | lp:~abentley/launchpad/same-queue-query-fmr |
| Merge into: | lp:launchpad |
| Diff against target: |
65 lines (+21/-12) 1 file modified
lib/lp/services/job/tests/test_celeryjob.py (+21/-12) |
| To merge this branch: | bzr merge lp:~abentley/launchpad/same-queue-query-fmr |
| Related bugs: |
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Abel Deuring (community) | code | 2012-08-06 | Approve on 2012-08-07 |
|
Review via email:
|
|||
Commit Message
Directly use matching queue in test_find_
Description of the Change
= Summary =
test_find_
== Proposed fix ==
When the queue has been determined to be in the correct state, use that query directly instead of re-querying the queue.
== Pre-implementation notes ==
Discussed with deryck
== LOC Rationale ==
I have a LOC credit of 1860
== Implementation details ==
Implement getFMR, as a variant of assertQueueSize that uses FindMissingReady to query the queue, and returns the instance that had a successful query.
== Tests ==
bin/test -t test_find_
== Demo and Q/A ==
None
= Launchpad lint =
Checking for conflicts and issues in changed files.
Linting changed files:
lib/lp/

I see how this is an improvement: only one call of list_queued(), so r=me. But the fact that list_queued() can obviously return different data in two calls when the message server should not have received of delivered any message, at least for Celery related queues, still scares me...