summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Reintroduce the three levels of forcedness. We do still need it.Simon MacMullen2012-02-012-16/+34
|
* Inline.Simon MacMullen2012-02-011-9/+10
|
* This doesn't start any more.Simon MacMullen2012-02-011-2/+1
|
* Base discnodeness on what we want to become, not what we are.Simon MacMullen2012-02-011-2/+2
|
* Change the semantics of Force to fit this particular bug. *meep*.Simon MacMullen2012-02-011-11/+7
|
* Rebase 038bd3c185aeSimon MacMullen2012-01-311-0/+8
|
* Resurrect and remove redundant ensure_mnesia_running()Simon MacMullen2012-01-311-1/+0
|
* ...however, we should still allow rabbitmqctl reset to work when some nodes ↵Simon MacMullen2012-01-241-1/+4
| | | | are down. (This is what we did before, the difference between reset and force_reset is that force_reset works if mnesia is completely hosed).
* Don't force on init() (which fixes bug 24703), plus a little refactor.Simon MacMullen2012-01-241-4/+5
|
* Merge bug24653Simon MacMullen2012-01-2311-118/+124
|\
| * refactor: use types exported by 'supervisor'Matthias Radestock2012-01-231-19/+4
| | | | | | | | instead of defining our own
| * helper function for constructing tcp listener sup child specsMatthias Radestock2012-01-231-13/+24
| | | | | | | | useful in e.g. stomp
| * use the correct inet typesMatthias Radestock2012-01-233-4/+4
| |
| * use correct type for {M,F,A}Matthias Radestock2012-01-239-26/+43
| | | | | | | | the built-in mfa() type stands for module-function-*arity*
| * refactor: check_tcp_listener_address/2 -> tcp_listener_addresses/1Matthias Radestock2012-01-231-23/+21
| | | | | | | | ...which is a more sensible name and signature
| * refactor: turn a condition into a guardMatthias Radestock2012-01-231-10/+6
| |
| * cosmetic: move helper functions where they belongMatthias Radestock2012-01-231-33/+32
| |
* | handle the remaining enabled_plugins_file error case more gracefullyMatthias Radestock2012-01-221-0/+2
| |
* | don't complain about empty enabled_plugins fileMatthias Radestock2012-01-221-0/+1
|/
* Merged bug23432 into defaultEmile Joubert2012-01-201-1/+1
|\
| * Avoid a circular dependency. The new load_definitions depends on recovery ↵Simon MacMullen2012-01-051-1/+1
| | | | | | | | | | | | (since it may create a queue and recovery starts the sup). But recovery depends on empty_db_check. But load_definitions *enables* empty_db_check (since we want to avoid creating default_foo if we just loaded a bunch of definitions). But there's a simple fix: recovery doesn't actually need to depend on empty_db_check.
* | Resurrect defaultSimon MacMullen2012-01-201-2/+3
| |
* | Merge bug23938Simon MacMullen2012-01-194-70/+73
|\ \
| * | refer to rationale from bug 24607Matthias Radestock2011-12-011-15/+2
| | |
| * | add upgrade stepMatthias Radestock2011-12-011-0/+8
| | |
| * | tie up some loose ends for the new tableMatthias Radestock2011-11-301-5/+5
| | |
| * | now that we have table locks there is no need for phase separationMatthias Radestock2011-11-301-66/+32
| | | | | | | | | | | | | | | This is essentially reverting to the code as it was before bug 23885, except using the new rabbit_topic_trie_node table for counts.
| * | only lock tables when we are dealing with multiple bindingsMatthias Radestock2011-11-301-6/+11
| | |
| * | simplifyMatthias Radestock2011-11-301-18/+6
| | |
| * | reduce time complexity of binding removal on queue deletionMatthias Radestock2011-11-303-19/+68
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 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
* | | correct typoMatthias Radestock2012-01-191-2/+2
| | |
* | | Merged bug24641 into default.Jerry Kuch2012-01-181-0/+4
|\ \ \
| * | | Add VERSION to the build and destroot phase environments for MacPortsSteve Powell2012-01-061-0/+4
| | | |
* | | | Same fix as c412171663cc for rabbitmq-pluginsSimon MacMullen2012-01-171-2/+9
| | | |
* | | | merge bug 24265 into default (Somehow support DNs in ↵Alexandru Scvortov2012-01-1719-156/+148
|\ \ \ \ | | | | | | | | | | | | | | | rabbitmq_auth_mechanism_ssl)
| * | | | clearly this was tested thoroughlyMatthias Radestock2012-01-161-1/+1
| | | | |
| * | | | Merge bug 24638Simon MacMullen2012-01-162-1/+2
| |\ \ \ \
| | * | | | increase closing timeoutMatthias Radestock2012-01-101-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | to increase the chances that buffered data will make it to the client.
| | * | | | don't lingerMatthias Radestock2012-01-101-0/+1
| | | | | |
| * | | | | Merge bug24676Simon MacMullen2012-01-161-3/+11
| |\ \ \ \ \
| | * \ \ \ \ Merge bug24674Simon MacMullen2012-01-169-130/+95
| | |\ \ \ \ \
| | * | | | | | fix rabbitmqctl error reporting under R15Matthias Radestock2012-01-121-3/+11
| | | | | | | |
| * | | | | | | fix supervisor2 bug reported by dialyzerMatthias Radestock2012-01-161-2/+2
| | |/ / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The erroneous code is invoked when a child terminates normally at the same time as it is being terminated by the supervisor. So unsurprisingly we haven't seen this, and it's hard to test.
| * | | | | | Merge bug24681Simon MacMullen2012-01-160-0/+0
| |\ \ \ \ \ \
| * | | | | | | merge rabbit_router:deliver into rabbit_amqqueue:deliverMatthias Radestock2012-01-167-92/+67
| |/ / / / / /
| * | | | | | Revert b03fe604af24. Not enough sleep.Simon MacMullen2012-01-131-2/+2
| | | | | | |
| * | | | | | Merge bug 24673Simon MacMullen2012-01-130-0/+0
| |\ \ \ \ \ \
| * | | | | | | Emit the timeout in the stats event.Simon MacMullen2012-01-131-2/+2
| |/ / / / / /
| * | | | | | a little bit of refactoring on the readerMatthias Radestock2012-01-121-35/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - move the 'blocked'-induced heartbeater disabling to the point at which the transition to the 'blocked' state takes place. This makes the correlation between these two events more obvious. It also prevents duplicate invocations of rabbit_heartbeat:pause_monitor/1, which hitherto was accomplished by relying on switch_callback/3 never getting called while 'blocked' - an assumption that could easily become invalid one day. - a little bit of inlining of timeout calculation code in close_connection/1. - more compact case handling in handle_dependent_exit/3 and wait_for_channel_termination/2.
| * | | | | | Merge bug24636 moreEmile Joubert2012-01-121-3/+5
| |\ \ \ \ \ \ | | |/ / / / / | |/| | | | |
View Lorry Controller Status