summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Merging default into bug 21673Matthew Sackman2010-02-0511-53/+648
|\
| * Merging bug 22309 onto defaultMatthew Sackman2010-02-052-2/+11
| |\
| | * handle attempt to close connection with a non-connection pidMatthias Radestock2010-02-052-2/+11
| | | | | | | | | | | | | | | which was previously hanging at best and causing things to explode at worst
| * | Merging bug22300 into defaultMatthew Sackman2010-02-055-16/+74
| |\ \
| | * | Correction of specMatthew Sackman2010-02-051-1/+2
| | | |
| | * | exercise some more code paths in testsMatthias Radestock2010-02-041-4/+6
| | | |
| | * | oopsMatthias Radestock2010-02-041-1/+1
| | | |
| | * | ditch API for channel-based consumer listingMatthias Radestock2010-02-041-18/+1
| | | |
| | * | move consumer listing from channel to queueMatthias Radestock2010-02-045-20/+48
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The main motivation here is that queues, not channels, are the logical home of subscriptions. And rabbit plug-ins and other extensions may bypass channels when subscribing to queues. The queues also know more about consumers than channels, namely whether a consumer requires acks. So this additional info we can now display too. And with the switch to queues the listing is now scoped by vhost.
| | * | test for list_consumersMatthias Radestock2010-02-041-6/+12
| | | |
| | * | merge bug21966 into bug22300Matthias Radestock2010-02-0410-45/+99
| | |\ \
| | * | | add 'rabbitmqctl list_consumers' commandMatthias Radestock2010-02-022-0/+21
| | | | |
| | * | | extend channel API with funs to get hold of consumersMatthias Radestock2010-02-021-1/+18
| | | | |
| | * | | drag in code from bug21966Matthias Radestock2010-02-025-11/+358
| | |\ \ \ | | | | | | | | | | | | | | | | | | since we need the ability to find all channels
| * | \ \ \ Merging bug 22309 onto defaultMatthew Sackman2010-02-055-18/+105
| |\ \ \ \ \ | | | |_|_|/ | | |/| | |
| | * | | | Minor correction to regexpMatthew Sackman2010-02-051-1/+1
| | | | | |
| | * | | | docsMatthias Radestock2010-02-042-7/+13
| | | | | |
| | * | | | add test for 'rabbitmqctl close_connection'Matthias Radestock2010-02-041-2/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This provides at least *some* coverage. Testing what happens when a connection is in the 'running' state is too hard.
| | * | | | refactor: move pid<->string conversion funs into miscMatthias Radestock2010-02-042-56/+64
| | | | | | | | | | | | | | | | | | | | | | | | since they are generally useful
| | * | | | merge default into bug22309Matthias Radestock2010-02-048-32/+46
| | |\ \ \ \
| | * | | | | better error reportingMatthias Radestock2010-02-031-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Don't mis-report errors on the remote end as "unable to connect" errors. The reason this didn't come up previously is that the function we call for close_connection is the first to fail with an 'EXIT' signal rather than a 'throw'.
| | * | | | | add 'rabbitmqctl close_connection'Matthias Radestock2010-02-031-0/+54
| | | | | | |
| | * | | | | extend rabbit_reader API with a shutdown funMatthias Radestock2010-02-031-8/+21
| | | |/ / / | | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is wired into the same logic as the handling of a supervisor-initiated shutdown. When the connection is in the 'running' state we send a 'connection.close' with a CONNECTION_FORCED code and follow our normal connection closure logic. In any other state we just drop out of the mainloop and thus close the socket.
| * | | | | merging bug 22301 into defaultMatthew Sackman2010-02-053-1/+33
| |\ \ \ \ \
| | * | | | | extend 'rabbitmqctl list_queues' command with info on exclusivityMatthias Radestock2010-02-023-1/+33
| | |/ / / /
| * | | | | merging bug 21966 onto defaultMatthew Sackman2010-02-056-18/+427
| |\ \ \ \ \ | | |_|/ / / | |/| | | |
| | * | | | make channel 'messages_unacknowledged' item semantics match that of queuesMatthias Radestock2010-02-043-4/+14
| | | |_|/ | | |/| | | | | | | | | | | | | | | | | namely, don't decrease the count for uncommitted acks Also, add 'acks_uncommitted' info item, for completeness
| | * | | add tests for pg_localMatthias Radestock2010-02-041-0/+23
| | | | | | | | | | | | | | | | | | | | these give us nearly complete code coverage
| | * | | add pg_local:sync helper function for testingMatthias Radestock2010-02-041-0/+10
| | | | | | | | | | | | | | | | | | | | | | | | | turns out that making any kind of assertions when everything is async is pretty much impossible
| | * | | automate testing of 'rabbitmqctl list_channels'Matthias Radestock2010-02-032-1/+10
| | | | | | | | | | | | | | | | | | | | and add rabbit_channel:info_keys/0, in line with other info-providing modules
| | * | | add info item for channel numberMatthias Radestock2010-02-033-4/+11
| | | | |
| | * | | merge default into bug21966Matthias Radestock2010-02-038-39/+55
| | |\ \ \
| | * | | | introduce exponential hibernation backoff in channel processMatthias Radestock2010-02-031-8/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | so that things like a scheduled 'rabbitmqctl list_channels' don't cause the channel to spend all its time hibernating and waking up
| | * | | | minor refactorMatthias Radestock2010-02-031-6/+5
| | | |_|/ | | |/| |
| | * | | fix typosMatthias Radestock2010-02-021-5/+4
| | | | |
| | * | | cosmeticMatthias Radestock2010-02-011-1/+1
| | | | |
| | * | | add 'rabbitmqctl list_channels' commandMatthias Radestock2010-02-012-0/+69
| | | | |
| | * | | add prefetch_count channel info itemMatthias Radestock2010-02-012-4/+19
| | | | | | | | | | | | | | | | | | | | | | | | | This forces us to make the limiter a gen_server*2* at last, since we want to issue info-like calls at higher priority, as elsewhere.
| | * | | first set of channel info itemsMatthias Radestock2010-02-011-1/+55
| | | | | | | | | | | | | | | | | | | | the low hanging fruit
| | * | | s/all/listMatthias Radestock2010-02-011-3/+3
| | | | |
| | * | | introduce channel registryMatthias Radestock2010-02-011-8/+15
| | | | |
| | * | | make pg_local do what it is supposed toMatthias Radestock2010-02-011-244/+71
| | | | |
| | * | | rename pg2 to pg_local, prior to more substantial modificationsMatthias Radestock2010-02-011-1/+1
| | | | |
| | * | | check in R13B-3 version of pg2, before modificationMatthias Radestock2010-02-011-0/+376
| | | | |
* | | | | Merging default into bug21673Matthew Sackman2010-02-042-6/+18
|\ \ \ \ \ | |/ / / /
| * | | | Merging bug22254 into defaultMatthew Sackman2010-02-042-6/+18
| |\ \ \ \ | | |_|/ / | |/| | |
| | * | | merging default into bug 22254Matthew Sackman2010-02-0412-70/+144
| | |\ \ \ | | |/ / / | |/| | |
| | * | | improving shortstr overflow checkingVlad Ionescu2010-01-292-6/+18
| | | | |
* | | | | Merging default into bug21673Matthew Sackman2010-02-049-41/+57
|\ \ \ \ \ | |/ / / /
| * | | | allow programmatic extraction of available info keysMatthias Radestock2010-02-036-26/+28
| | |_|/ | |/| | | | | | | | | | | | | | ...which avoids copy&paste, instantly improves test coverage, and seems generally useful.
View Lorry Controller Status