summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Don't allow rabbit_misc:is_process_alive/1 to hang indefinitely if a node is ↵Simon MacMullen2014-07-251-1/+3
| | | | incommunicado; we use this in rabbit_amqqueue:on_node_down/1 in a loop!
* Merge in stableSimon MacMullen2014-07-251-2/+20
|\
| * Avoid connecting to nodes that are down.Simon MacMullen2014-03-111-2/+20
| |
* | Merge bug26290Simon MacMullen2014-07-253-7/+56
|\ \
| * | InlineSimon MacMullen2014-07-251-8/+4
| | |
| * | Make vm_memory_monitor:parse_line_linux/1 parse lines w/o colonsMichael Klishin2014-07-251-8/+25
| | | | | | | | | | | | Happens with some broken kernel modules.
| * | Tests for vm_memory_monitor:parse_line_linux/1Michael Klishin2014-07-252-0/+36
| | |
* | | Uploaders field change in debian/control to announce Debian Maintainer ↵Blair Hester2014-07-241-1/+1
| | | | | | | | | | | | change from Emile to Blair
* | | Merge bug26293Simon MacMullen2014-07-223-34/+64
|\ \ \ | |/ / |/| |
| * | Actually, hold on to confirms when pausing. It's not inconceivable that we ↵Simon MacMullen2014-07-221-1/+1
| | | | | | | | | | | | might get a false positive from pause_minority_guard, in which case we don't want to throw data away.
| * | Check for pause minority mode being about to cause a shutdown before sending ↵Simon MacMullen2014-07-223-34/+64
|/ / | | | | | | confirms. In this way we can get rid of the effort to kill all network connections ASAP.
* | Merge bug26292Simon MacMullen2014-07-160-0/+0
|\ \
* | | Further documentation of this approach.Simon MacMullen2014-07-161-13/+11
| | |
* | | It's probably a good idea to reset "restarts" after we get the message, not ↵Simon MacMullen2014-07-141-3/+4
| | | | | | | | | | | | when we send it; we don't want to jump back into restarting things as normal just because we have scheduled a restart.
* | | Reset restarts before we go round again, otherwise if MaxT is large we will ↵Simon MacMullen2014-07-141-1/+1
|/ / | | | | | | keep looping until it clears.
* | Merge bug26272Simon MacMullen2014-06-301-0/+4
|\ \
| * | Emit an event when we modify a global parameter.Simon MacMullen2014-06-301-0/+4
|/ /
* | Merge bug26225Simon MacMullen2014-06-273-14/+45
|\ \
| * | Separate out responsibilities in the various node state detection functions. ↵Simon MacMullen2014-06-272-8/+19
| | | | | | | | | | | | Only ping_all/0 is allowed to establish new tcp connections (and thus take significant time for them to time out if necessary). This removes a significant delay while waiting for pause_minority to start.
| * | Hard-kill all connections and stop all listeners as soon as we decide we are ↵Simon MacMullen2014-06-252-6/+26
|/ / | | | | | | in a minority. Also remve an export from rabbit_networking that was unused (probably dated from era of unsupervised plugins).
* | Added tag rabbitmq_v3_3_4 for changeset 8e9bc23a55eaSimon MacMullen2014-06-240-0/+0
| |
* | Changelog for 3.3.4.rabbitmq_v3_3_4Simon MacMullen2014-06-242-0/+9
| |
* | Merge bug 26258Simon MacMullen2014-06-231-2/+2
|\ \
| * | OopsSimon MacMullen2014-06-191-1/+1
| | |
| * | Merge bug 26253Simon MacMullen2014-06-191-3/+3
| |\ \
| | * | Don't explode if we have a file with 0 or 2+ Erlang terms. Also remove TODO, ↵Simon MacMullen2014-06-191-3/+3
| |/ / | | | | | | | | | actually I don't think we can do better.
* | | More explanation.Simon MacMullen2014-06-231-2/+3
| | |
* | | Don't use -e for a portion of rabbitmq-env that uses readlinkMichael Klishin2014-06-231-0/+4
|/ / | | | | | | readlink exists with an error when its argument is not a symlink.
* | Merge bug26247Simon MacMullen2014-06-180-0/+0
|\ \
* | | Allow prelaunch to return error code 2 without failing.Simon MacMullen2014-06-181-2/+9
|/ /
* | Added tag rabbitmq_v3_3_3 for changeset 7cbb21091d12Simon MacMullen2014-06-170-0/+0
| |
* | changelogs for 3.3.3rabbitmq_v3_3_3Simon MacMullen2014-06-172-0/+9
| |
* | Merge bug26241Simon MacMullen2014-06-130-0/+0
|\ \
* | | Ensure the limit field is always a number so we can't miss the need to ↵Simon MacMullen2014-06-131-10/+10
|/ / | | | | | | interpret it. Yes, this means we won't update mem_relative if the size of the machine's physical memory changes at run time. I can live with that.
* | Merge bug26236 againSimon MacMullen2014-06-120-0/+0
|\ \
| * \ Merge bug26236Simon MacMullen2014-06-111-4/+6
| |\ \
* | | | Since a crash in handle_event is so disasterous, and since gen_event ↵Simon MacMullen2014-06-123-6/+31
| |/ / |/| | | | | | | | handlers are not very supervisable, let's wrap everything in a try/catch. Prinitng to stdout is not great but we just failed to log properly - better to give some indication than none.
* | | OopsSimon MacMullen2014-06-111-1/+1
| | |
* | | Don't fail if asked to truncate a reference. And go to having a catch-all case.Simon MacMullen2014-06-111-4/+6
|/ /
* | Resurrect stable (without bug 26230)Simon MacMullen2014-06-111-1/+0
| |
* | Merge bug24759Simon MacMullen2014-06-091-1/+1
|\ \
| * | Ensure we don't fail if that file is missing.Simon MacMullen2014-06-091-1/+1
| | |
* | | Merge bug24759Simon MacMullen2014-06-095-5/+5
|\ \ \ | |/ /
| * | Use '-e' for all shell scripts.Simon MacMullen2014-06-095-5/+5
|/ /
* | Fix tests for 32 bit machines.Simon MacMullen2014-06-091-2/+8
| |
* | Added tag rabbitmq_v3_3_2 for changeset 9ec4d3a2498aSimon MacMullen2014-06-090-0/+0
| |
* | changelogs for 3.3.2.rabbitmq_v3_3_2Simon MacMullen2014-06-092-0/+9
| |
* | Merge bug26227Simon MacMullen2014-06-052-3/+20
|\ \
| * | Don't show misleading information in the Windows service case when the ↵Simon MacMullen2014-06-052-3/+20
|/ / | | | | | | service has not been reinstalled.
* | Merge bug26226Simon MacMullen2014-06-041-1/+1
|\ \
View Lorry Controller Status