summaryrefslogtreecommitdiff
path: root/src
Commit message (Expand)AuthorAgeFilesLines
* make channel/soft errors easier to distinguish from hard errorsMatthias Radestock2012-07-161-1/+1
* more consistent error handling in readerMatthias Radestock2012-07-152-105/+128
* call the correct functionMatthias Radestock2012-07-151-1/+1
* refactor: precondition_failed helper functionMatthias Radestock2012-07-151-21/+17
* oops - 'hg record' diff boundary failMatthias Radestock2012-07-151-0/+4
* renamingMatthias Radestock2012-07-151-11/+7
* cosmeticMatthias Radestock2012-07-141-2/+2
* tiny simplifying refactorMatthias Radestock2012-07-131-2/+1
* Merged bug25052 into defaultEmile Joubert2012-07-131-5/+5
|\
| * blow up if Ref is unknownMatthias Radestock2012-07-121-1/+1
| * grammarMatthias Radestock2012-07-121-1/+1
| * Duh, cosmeticSimon MacMullen2012-07-121-2/+1
| * Don't reopen file handles in fhc:needs_sync/1, avoid infinite loop.Simon MacMullen2012-07-121-5/+6
* | I am more dubious about this one. But not enough to take it out.Simon MacMullen2012-07-131-0/+2
* | Explain this message.Simon MacMullen2012-07-131-0/+5
* | Rename HA heartbeat when master changesEmile Joubert2012-07-112-4/+4
* | Remove unnecessary HA heartbeatsEmile Joubert2012-07-101-11/+0
|/
* Merged bug24826 into defaultEmile Joubert2012-06-2910-89/+90
|\
| * merge defaultTim Watson2012-06-2721-831/+892
| |\
| | * It probably makes slightly more sense to monitor before we broadcast.Simon MacMullen2012-06-261-1/+1
| | * Maybe better name?Simon MacMullen2012-06-263-3/+3
| | * OK, we do actually need a special way to determine the difference between:Simon MacMullen2012-06-253-3/+3
| | * Get rid of this with_exit_handler use.Simon MacMullen2012-06-251-12/+4
| | * normal -> leaving the ring, change topologySimon MacMullen2012-06-253-5/+5
| | * And finally, have the master wait for all the slaves on deletion.Simon MacMullen2012-06-251-0/+11
| | * The spec for gm:group_members/1 says it returns a list of pids - let's do that.Simon MacMullen2012-06-251-1/+1
| | * Ignore normal exits; do not try to change the ring topology as it shuts down.Simon MacMullen2012-06-251-3/+5
| | * Close down the slave's GM the same way the master's GM does.Simon MacMullen2012-06-251-0/+5
| | * consistency: leave the 'ok' matching to the callerMatthias Radestock2012-06-231-1/+1
| | * Revert the wrapper around the copy in append_file/2Simon MacMullen2012-06-221-0/+17
| | * Merge bug24956Simon MacMullen2012-06-203-24/+16
| | |\
| | | * Merged bug24992 into defaultEmile Joubert2012-06-191-1/+1
| | | |\
| | | | * Default to 1GB limit, check every 10s.Simon MacMullen2012-06-191-1/+1
| | | * | Merge bug24994Simon MacMullen2012-06-191-1/+5
| | | |\ \
| | | | * \ Merge bug24939Simon MacMullen2012-06-141-2/+2
| | | | |\ \ | | | | | |/
| | | | * | Strip this out again, we can't make it idempotent and it might just remove th...Simon MacMullen2012-06-141-3/+0
| | | | * | Ensure we only clear out records of queues that are actually dead.Simon MacMullen2012-06-141-1/+2
| | | | * | This has little to do with recovery - move to networking boot.Simon MacMullen2012-06-142-1/+3
| | | | * | Improve commentSimon MacMullen2012-06-141-1/+3
| | | | * | clean up listeners during node start/recoverySteve Losh2012-06-141-0/+1
| | | | * | merge defaultSteve Losh2012-06-1422-966/+1121
| | | | |\ \
| | | | * | | get a node to clean up after old incarnations of itselfMatthias Radestock2012-05-161-0/+1
| | | * | | | CosmeticEmile Joubert2012-06-181-1/+0
| | | * | | | Only one version of append_file is requiredEmile Joubert2012-06-183-28/+4
| | | * | | | Rotate logs differentlyEmile Joubert2012-06-143-9/+13
| | | * | | | Rotate logs without consuming excessive memoryEmile Joubert2012-06-143-5/+14
| | | | |_|/ | | | |/| |
| | * | | | revert the log level for error for heartbeat timeoutsTim Watson2012-06-181-1/+0
| | * | | | make the error/reason for amqp heartbeat timeouts clearer in the logsTim Watson2012-06-181-4/+4
| | * | | | merge defaultTim Watson2012-06-147-103/+79
| | |\ \ \ \ | | | |/ / /
| | | * | | Report more clearly when we fail to monitor disk space on an unsupported plat...Simon MacMullen2012-06-141-2/+2
| | | | |/ | | | |/|