summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* merge headsMatthias Radestock2012-03-221-1/+3
|\
| * fix specMatthias Radestock2012-03-221-1/+3
| |
| * Added tag rabbitmq_v2_8_1 for changeset 99239db7a754Simon MacMullen2012-03-220-0/+0
| |
* | cosmeticMatthias Radestock2012-03-221-3/+2
| |
* | consistencyMatthias Radestock2012-03-221-1/+1
| |
* | cosmetic: log queue name with rabbit_misc:rs/1, like everywhere elseMatthias Radestock2012-03-221-2/+2
| |
* | fix a possible leakMatthias Radestock2012-03-221-4/+3
|/ | | | | | | | | | it's not clear whether we can actually end up in the 'none' branch, but assuming it exists for a reason, we should certainly clear the dict entry there too. We don't actually need to clear the entry in the {value, ...} branch, since handle_confirm will do that. But clearing in all cases is more obviously correct.
* Merged bug24809 into defaultrabbitmq_v2_8_1Emile Joubert2012-03-221-1/+3
|\
| * Rollback startup method for SuSE onlyEmile Joubert2012-03-221-1/+3
| |
* | Accomodate a desire to avoid polluting the logsMatthew Sackman2012-03-221-0/+13
|/
* Merge bug24816.Simon MacMullen2012-03-220-0/+0
|\
* | Package all bat files.Simon MacMullen2012-03-221-4/+1
|/
* Merge bug24810 again again.Simon MacMullen2012-03-213-2/+5
|\
| * So we actually need to handle the case of starting from scratch separately, ↵Simon MacMullen2012-03-213-2/+5
| | | | | | | | otherwise we never write out the version file, with fun consequences.
* | Merged bug24810 into defaultEmile Joubert2012-03-212-2/+13
|\ \ | |/
| * It's possible somehow we could be left with an empty mnesia directory ↵Simon MacMullen2012-03-212-2/+10
| | | | | | | | (spotted with "make cleandb" but maybe other things could do this). So instead check for the existence of the guid serial, which has not changed since at least 1.7.0.
| * If don't even have an mnesia dir yet, treat that as meaning "no upgrades ↵Simon MacMullen2012-03-211-1/+4
| | | | | | | | needed" rather than "upgrades are impossible"
* | Merging bug24774 to defaultMatthew Sackman2012-03-210-0/+0
|\ \
* | | And if an {intrinsic, Delay} child exits normally/shutdown, remember to ↵Matthew Sackman2012-03-211-0/+2
| | | | | | | | | | | | really exit the supervisor.
* | | Some cosmetics; Some consistency of ordering; There is no good reason why ↵Matthew Sackman2012-03-211-14/+17
| | | | | | | | | | | | delayed restart shouldn't also support intrinsic. Added.
* | | Apply supplied patch to support normal/shutdown termination of delayed ↵Ivan Dyachkov2012-03-211-10/+15
|/ / | | | | | | restart children
* | Merged bug24810 into defaultEmile Joubert2012-03-210-0/+0
|\ \ | |/ |/|
* | Look for cluster nodes in Mnesia's view of the world (in case we are a disc ↵Simon MacMullen2012-03-201-1/+4
|/ | | | node) and the config file (in case we are a ram node)
* Added tag rabbitmq_v2_8_0 for changeset 152dab3f6974Simon MacMullen2012-03-190-0/+0
|
* Merged bug24794 into defaultrabbitmq_v2_8_0Emile Joubert2012-03-154-13/+82
|\
| * Simple test for multi call - and reverse the lists so they come out the same ↵Simon MacMullen2012-03-152-1/+22
| | | | | | | | order they went in.
| * Slightly nicer to use a -define().Simon MacMullen2012-03-151-1/+3
| |
| * CommentSimon MacMullen2012-03-141-1/+2
| |
| * Only sleep as much as we need, and tail recurse properly.Simon MacMullen2012-03-141-6/+5
| |
| * Hush, little dialyzer, don't cry.Simon MacMullen2012-03-141-2/+2
| |
| * Scatter-gather.Simon MacMullen2012-03-142-4/+33
| |
| * This fixes the bug, but it does not scatter-gather, which would certainly be ↵Simon MacMullen2012-03-142-12/+29
|/ | | | nice...
* Merge bug24614Simon MacMullen2012-03-085-32/+59
|\
| * Merge bug24779Simon MacMullen2012-03-073-13/+15
| |\
| | * Let it crash...Simon MacMullen2012-03-071-7/+1
| | |
| | * Ahem.Simon MacMullen2012-03-071-1/+1
| | |
| | * Try to close the socket immediately on reader death.Simon MacMullen2012-03-072-6/+14
| | |
| * | Add to qi the equivalent of msg_store_file:scan and rework a specialised ↵Matthew Sackman2012-03-061-12/+38
| |/ | | | | | | walker so that it makes use of the generalised one. This is exported so that it can be used elsewhere.
| * cosmetic, plus report Pid instead of {ok, Pid}Matthias Radestock2012-03-051-7/+6
| |
* | Not needed.Simon MacMullen2012-03-081-1/+1
| |
* | Make RedHat/Fedora use runuser to start the serviceSteve Powell2012-03-051-0/+1
| |
* | Reduce changes and preserve atomicity of redirection filesSteve Powell2012-03-051-2/+3
| |
* | Modify packaging for Debian to use start-stop-daemonSteve Powell2012-03-052-5/+6
|/
* cosmeticMatthias Radestock2012-03-021-3/+3
|
* Merge bug 24768Simon MacMullen2012-03-010-0/+0
|\
* | In fact, let's not call handle_ch_down in the case where we get a 'DOWN' ↵Simon MacMullen2012-03-011-5/+9
| | | | | | | | while waiting to stop.
* | And a commentSimon MacMullen2012-03-011-0/+2
| |
* | Actually be slightly more minimal about what we move.Simon MacMullen2012-03-011-10/+10
| |
* | Don't throw away 'DOWN' in delayed_stop.Simon MacMullen2012-03-011-15/+16
|/
* Merge bug 22961Simon MacMullen2012-03-016-23/+58
|\
View Lorry Controller Status