summaryrefslogtreecommitdiff
path: root/include
Commit message (Collapse)AuthorAgeFilesLines
...
* | | | | | | tweak credit_flow logicMatthias Radestock2012-02-061-1/+1
| |_|_|_|_|/ |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - make credit specs contain the InitialCredit and MoreCreditAfter. This is more intuitive and means the message sender only needs to look at the former and the receiver at the latter. - narrow the condition on which handle_bump_msg invokes unblock - it only needs to do that on *transition* from =< 0 to > 0. - refactor: extract process dict update - cosmetic: more logical order of exported functions
* | | | | | Update copyright 2012bug24606Emile Joubert2012-02-028-9/+9
| | | | | |
* | | | | | Merged bug24671 into defaultEmile Joubert2012-01-311-10/+2
|\ \ \ \ \ \
| * | | | | | get rid of debug and message loggingMatthias Radestock2012-01-261-10/+0
| | |_|_|_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | since they have limited use This does make the rabbit_log server somewhat superfluous, but it does get a new lease of life in bug 22470
| * | | | | reduce time complexity of binding removal on queue deletionMatthias Radestock2011-11-301-0/+2
| | |_|_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | from O(n^2) to O(n). The solution comes in two parts: 1) a new table, rabbit_topic_trie_node, to explicitly track edge and binding counts of nodes in the trie. This eliminates expensive mnesia:match_object operations to determine these counts when needed. 2) table-scope write locks to eliminate row-level locking, which is expensive when there are many such locks in a single transaction
* | | | | Remove ?CREDIT_CPU_BOUND and have /1 versions of ack and send instead.bug24671Simon MacMullen2012-01-241-1/+0
| | | | |
* | | | | More credit for the queue -> msg_store case.Simon MacMullen2012-01-191-0/+2
|/ / / /
* | | | Refrain from resetting message expiry in message requeueEmile Joubert2011-10-201-4/+1
|/ / /
* | | the plugins distributions may be either ezs or directoriesAlexandru Scvortov2011-09-221-0/+1
| | | | | | | | | | | | Turns out the token parser was broken: it didn't work with nested tuples.
* | | don't use the plugins directoryAlexandru Scvortov2011-09-221-0/+6
| | |
* | | backing_queue_spec and variable_queue specs up to dateAlexandru Scvortov2011-08-051-2/+3
| | |
* | | merge default into bug24216Matthias Radestock2011-07-042-15/+5
|\ \ \
| * | | tx gone from everywhereMatthias Radestock2011-06-241-2/+1
| | | |
| * | | remove transactions from backing queueMatthias Radestock2011-06-241-13/+4
| | | | | | | | | | | | | | | | ...and mostly from amqqueue_process
* | | | Scratch space for exchanges, and rabbit_exchange:update/2.Simon MacMullen2011-06-271-1/+2
|/ / /
* | | Sort out (everywhere except mirror modules) slave pids and mirror nodes and ↵Matthew Sackman2011-06-211-1/+1
| | | | | | | | | | | | deal with non-local queue declaration, and introduce ha policy and ha policy params. (does not compile)
* | | Merging default into bug23554Matthew Sackman2011-06-211-1/+1
|\ \ \
| * | | Permit dropping nodes of mirrored queues. This turns out to be much much ↵Matthew Sackman2011-05-241-2/+2
| | | | | | | | | | | | | | | | messier than I'd hoped as the principle problem becomes ensuring an add after a drop works. Normally, an add would only occur on a node that has not seen said queue before: if it had, in a previous lifetime, then the booting of rabbit would have ripped out any locally stored files regarding that queue. But now this step may be missed. Having tried many different approaches, the simplest became expanding bq so that the shutdown reason is exposed to the BQ. Thus both slave and master can then detect that they're being dropped, and, in the case of master, it can convert a bq:terminate to a bq:delete_and_terminate. Every other approach I could think of turned out worse.
| * | | Merge with defaultRob Harrop2011-05-201-1/+1
| |\ \ \
| | * \ \ Merging default to bug23554Matthew Sackman2011-05-172-4/+9
| | |\ \ \
| | * \ \ \ Merging default into bug23554Matthew Sackman2011-04-102-13/+20
| | |\ \ \ \
| | | * | | | Well txns are still only half in. But I want to go home.Matthew Sackman2011-04-071-1/+2
| | | | | | |
| | | * | | | Add BQ:discard, correct BQ:is_duplicate, finally fix the last bits of ↵Matthew Sackman2011-04-071-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | immediate delivery, though hopefully in a way which has not leaked through to the lower layers...
| | | * | | | Validate_msg becomes is_duplicate, which makes much more senseMatthew Sackman2011-04-071-2/+2
| | | | | | |
| | | * | | | merge default into bug23554Matthew Sackman2011-03-141-3/+6
| | | |\ \ \ \
| | | * \ \ \ \ Merge default to bug23554, and then a bitMatthew Sackman2011-03-092-12/+17
| | | |\ \ \ \ \
| | | | * | | | | oopsMatthew Sackman2011-03-081-2/+2
| | | | | | | | |
| | | | * | | | | add BQ:validate_messageMatthew Sackman2011-03-081-0/+2
| | | | | | | | |
| | | | * | | | | Merging default into bug23554Matthew Sackman2011-03-082-12/+15
| | | | |\ \ \ \ \ | | | |/ / / / / /
| | | | * | | | | Merging default into bug23554Matthew Sackman2011-02-282-1/+11
| | | | |\ \ \ \ \
| | | | * \ \ \ \ \ Merging bug23727 to bug23554Matthew Sackman2011-02-232-2/+2
| | | | |\ \ \ \ \ \
| | | | * \ \ \ \ \ \ Merging bug23727 to bug23554Matthew Sackman2011-02-122-16/+22
| | | | |\ \ \ \ \ \ \
| | | | | * \ \ \ \ \ \ Merging default into bug23554Matthew Sackman2011-02-123-12/+43
| | | | | |\ \ \ \ \ \ \
| | | | | | * \ \ \ \ \ \ Merging bug23727 into bug23554Matthew Sackman2011-01-211-1/+1
| | | | | | |\ \ \ \ \ \ \
| | | | | | * \ \ \ \ \ \ \ Merging default into bug23554 (epic debitrot)Matthew Sackman2011-01-203-11/+42
| | | | | | |\ \ \ \ \ \ \ \
| | | | | | | * | | | | | | | Correct all manner of crazy wrongness with ↵Matthew Sackman2010-12-211-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | maybe_run_queue_via_backing_queue. The principle problem is that the state with which the function is to be run is dependent on the caller. Sometimes it's the 'top level' BQ state (eg invocations called by amqqueue_process). Sometimes it's the 'bottom level' state (eg invocations called by vq or qi or msg_store). We didn't used to have multiple layers, but now we do. Consequently, the invocation must indicate 'where' the fun is to be run. The clearest explanation is in master:invoke/3
| | | | | | | * | | | | | | | Merging default into bug 23554Matthew Sackman2010-12-213-10/+39
| | | | | | | |\ \ \ \ \ \ \ \
| | | | | | | | * | | | | | | | extra_pids => mirror_pids; Also rip out changes to the registry and just ↵Matthew Sackman2010-12-161-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | hard code that rabbit_mnesia takes care of gm - the problem being that we need access to the table dfns even when rabbit app is stopped, thus the registry would be down
| | | | | | | | * | | | | | | | Merging default into bug23554Matthew Sackman2010-12-161-1/+1
| | | | | | | | |\ \ \ \ \ \ \ \
| | | | | | | | * | | | | | | | | Move the various files from rabbit-ha to rabbit-server as the ha work can't ↵Matthew Sackman2010-12-161-0/+28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | exist as a plugin
| | | | | | | | * | | | | | | | | Merging default into bug23554Matthew Sackman2010-12-141-0/+41
| | | | | | | | |\ \ \ \ \ \ \ \ \
| | | | | | | | * | | | | | | | | | Alter the result of BQ:requeue so that it also returns the guids of the ↵Matthew Sackman2010-12-121-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | requeued messages. This makes it match the spec of ack closely and is more logical. Also, entirely coincidentally, happens to be necessary for HA...
| | | | | | | | * | | | | | | | | | Merge default into bug 23554 (epic debitrot)Matthew Sackman2010-12-062-3/+5
| | | | | | | | |\ \ \ \ \ \ \ \ \ \
| | | | | | | | * | | | | | | | | | | Exposing the publishing chpid right through to the BQ (and tidying up ↵Matthew Sackman2010-11-301-6/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | tests/types/specs)
| | | | | | | | * | | | | | | | | | | Make the tests work again. Also mirror_pids => extra_pids on the basis that ↵Matthew Sackman2010-11-291-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | it could be reusable for other purposes
| | | | | | | | * | | | | | | | | | | Lesser of many evils...Matthew Sackman2010-11-291-1/+1
| | | | | | | | | | | | | | | | | | |
| | | | | | | | * | | | | | | | | | | Present the whole amqqueue record to the BQ (remove duplicated IsDurable ↵Matthew Sackman2010-11-291-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | flag, correct BQ callbacks)
* | | | | | | | | | | | | | | | | | | Merge in defaultbug24164Simon MacMullen2011-06-171-1/+0
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ | | |_|_|_|_|_|_|_|_|_|_|_|_|_|_|_|_|/ | |/| | | | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | | | | | Make the stats interval configurable.bug24162Simon MacMullen2011-06-091-1/+0
| | |_|_|_|_|_|_|_|_|_|_|_|_|_|_|_|/ | |/| | | | | | | | | | | | | | | |
* | | | | | | | | | | | | | | | | | Since we're changing the backend API anyway, let's remove this ↵Simon MacMullen2011-06-091-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | management-specific wart.
View Lorry Controller Status