| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|\ |
|
| |\ |
|
| | |\ |
|
| | | |\ |
|
| | | | | |
|
| | | |/ |
|
| | | |\ |
|
| | | | |\ |
|
| | | | | |\
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
testConnectionQueueSameNode to fail due to a race in recovering
exclusive queues (bug 22695). We'll need a fix from default, so
leaving as is for the time being.
|
| | | | | |\ \ |
|
| | | | | | |\ \
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | | |
rabbit_amqqueue.erl, due in the main to recover/0 being removed and
refactoring of durable queue recovery in general.
|
| | | | | | |\ \ \ |
|
| | | | | | | |\ \ \ |
|
| | | | | | | |\ \ \ \
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | | |
20578) and exchange hooks intersect in a few places. The main things
to take care over are that the post-queue-deletion hooks gets called,
and the exclusivity check is called. NB: exchange auto-delete
machinery needs to be removed, which will simplify things in a few
places.
|
| | | | | | | | | | | | |
|
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | |
| | | | | | | | | | | | |
signature of rabbit_amqqueue:basic_consume/8
|
| | | | | | | |\ \ \ \ \
| | | | | | | | | | | | |
| | | | | | | | | | | | |
| | | | | | | | | | | | |
| | | | | | | | | | | | | |
the ../rabbitmq-test test suite. Some unused variables appear to have crept
in during previous commits to the branch; will fix in another commit.
|
| | | | | | | | | | | | | |
|
| | | | | | | | |\ \ \ \ \ |
|
| | | | | | | | | |\ \ \ \ \ |
|
| | | | | | | | |\ \ \ \ \ \ \
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | |
some tests
|
| | | | | | | | | | | | | | | | |
|
| | | | | | | | | |/ / / / / /
| | | | | | | | |/| | | | | |
| | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
as per the field doc in the spec.
|
| | | | | | | | |\ \ \ \ \ \ \
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | |
non-obvious conflict resolutions:
- getting rid of the q/x durability agreement check
- in getting rid of the exchange may_auto_deleting, making sure all
branches returned 'ok' (the tests all start failing!).
|
| | | | | | | | | |\ \ \ \ \ \ \ |
|
| | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
this synchronous. This means recover_async is now also synchronous,
but it's deprecated, and anyway we do the same thing with no_wait
elsewhere. This way we have just the one code path.
|
| | | | | | | | | | | | | | | | | |
|
| | | | | | | | | | |\ \ \ \ \ \ \ |
|
| | | | | | | | | | | | | | | | | | |
|
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
break frame-max
|
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
branches
|
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
it is exclusive to a dead process (connection); and if so, deleting
it. As things stand, a queue can't be owned by a connection on
another node (so the aliveness check is redundant), but this will
account for such scenarios in the event of connection- or
queue-mobility.
|
| | | | | | | | | | | |\ \ \ \ \ \ \
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | |
synchronous auto_deletes
|
| | | | | | | | | | | | | | | | | | | |
|
| | | | | | | | | | | | | | | | | | | |
|
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | |
to do the check inside the transaction, I had to change
rabbit_exchange:add_binding/4 (and delete_binding/4) to take another
parameter for the check.
I also made the exclusivity check more obviously run for its
side-effect.
|
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | |
in to the queue operations mentioned in the 0-9-1 spec's "exclusive"
rule (except for queue.bind which is a bit harder).
|
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | |
can check equivalence on queue.declare
- speaking of which, implement the "equivalent" rule of queue.declare
(bug 21832)
- queues are exclusive on creation, and don't change owner, so get rid
of claim_queue
- also check exclusive in the passive version of queue.declare; this
is explicitly mandated by the spec (queue.declare, exclusive)
- bug 21385: exclusive queues are deleted when their owners go away,
as per the spec. This didn't actually change implementation -- it
was wrong for 0-8
|
| | | | | | | | | | | | |\ \ \ \ \ \ \ |
|
| | | | | | | | | | | | | | | | | | | | |
|
| | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | | |
otherwise
|
| | | | | | | | | | | | | | | | | | | | |
|
| | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | | |
section 1.6)
|
| | | | | | | | | | | | | |\ \ \ \ \ \ \
| | | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | | | |
type specs, and a bad automerge in rabbit_reader.erl.
|
| | | | | | | | | | | | | | |\ \ \ \ \ \ \ |
|
| | | | | | | | | | | | | | | |\ \ \ \ \ \ \ |
|
| | | | | | | | | | | | | | | | | | | | | | | |
|
| | | | | | | | | | | | | | | |\ \ \ \ \ \ \ \ |
|
| | | | | | | | | | | | | | | | |\ \ \ \ \ \ \ \ |
|