summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Remove unused functionMatthew Sackman2010-11-111-3/+0
|
* exit rather than warn if future upgrades are found.Simon MacMullen2010-11-111-2/+1
|
* Comment formatting with M-q and no need to use ++ for multiline string constantsMatthew Sackman2010-11-111-6/+7
|
* CosmeticsMatthew Sackman2010-11-111-15/+11
|
* Prevent a very difficult-to-trigger race.Simon MacMullen2010-11-101-5/+4
|
* Fix use of file:open/2Simon MacMullen2010-11-101-1/+1
|
* Check both our version and the remote version. This is TRTTD I think, and ↵Simon MacMullen2010-11-091-0/+1
| | | | has the side effect that we write our version out on the subsequent node.
* Merged from defaultSimon MacMullen2010-11-095-87/+369
|\
| * Refactor a bit, have the first cluster node fail if version is wrongSimon MacMullen2010-11-081-43/+38
| |
| * Check the schema version on disc as well as the integrity check when ↵Simon MacMullen2010-11-051-15/+29
| | | | | | | | starting first disc cluster node
| * Keep the old behaviour when upgrading from 2.1.0 or earlier.Simon MacMullen2010-11-042-21/+29
| |
| * Check the version we want matches the one a remote node is running before ↵Simon MacMullen2010-11-042-23/+49
| | | | | | | | joining the cluster.
| * Create the schema version file in the case where we acquire the schema via ↵Simon MacMullen2010-11-041-1/+2
| | | | | | | | the cluster.
| * Blow up if the upgrade does not leave us with a valid schema. Refactor ↵Simon MacMullen2010-11-021-25/+31
| | | | | | | | several nested case expressions into one big one.
| * Inlining to prevent running out of disk space by overly long source filesMatthew Sackman2010-11-021-4/+3
| |
| * Same as before - avoid unnecessary functions for pattern matchMatthew Sackman2010-11-021-4/+4
| |
| * Match the list: no need to use lengthMatthew Sackman2010-11-021-4/+4
| |
| * Merging default into bug22573Matthew Sackman2010-11-0222-439/+571
| |\
| * | trivialMatthew Sackman2010-10-271-3/+2
| | |
| * | Even more beautifulMatthew Sackman2010-10-272-31/+27
| | |
| * | Remove hackSimon MacMullen2010-10-261-7/+3
| | |
| * | Merge in defaultSimon MacMullen2010-10-2613-432/+413
| |\ \
| * | | Abstract graph constructionMatthew Sackman2010-10-263-83/+92
| | | |
| * | | Specs.Simon MacMullen2010-10-252-1/+20
| | | |
| * | | Refactor and document upgrades_to_apply/2Simon MacMullen2010-10-251-5/+7
| | | |
| * | | Rename variableSimon MacMullen2010-10-251-2/+2
| | | |
| * | | Fix copyright statementsSimon MacMullen2010-10-252-24/+5
| | | |
| * | | Match more return values, and reindentSimon MacMullen2010-10-252-22/+26
| | | |
| * | | Undo 522af2a24c5e, wait_for_tables/0 in fact may not have been calledSimon MacMullen2010-10-251-0/+1
| | | |
| * | | Various stylistic tweaks + cleanupsSimon MacMullen2010-10-252-25/+23
| | | |
| * | | Rename module.Simon MacMullen2010-10-251-1/+1
| | | |
| * | | wait_for_tables has definitely been called by this point (in ↵Simon MacMullen2010-10-221-1/+0
| | | | | | | | | | | | | | | | check_table_integrity)
| * | | Merged from defaultSimon MacMullen2010-10-2217-362/+702
| |\ \ \
| * | | | Only attempt upgrade if we are *really* not clustered.Simon MacMullen2010-10-221-4/+4
| | | | |
| * | | | Only attempt upgrade if we are not clustered. wait_for_tables() beforehand ↵Simon MacMullen2010-10-221-1/+5
| | | | | | | | | | | | | | | | | | | | should thus be safe.
| * | | | mnesia:transform_table/3 appears to see table data before wait_for_tables, ↵Simon MacMullen2010-10-192-1/+19
| | | | | | | | | | | | | | | | | | | | which solves a lot of problems. That's probably deliberate :). Move invocation of maybe_upgrade and add a couple of test upgrades to change schema as well as data.
| * | | | Use mnesia:transform_tableSimon MacMullen2010-10-192-30/+12
| | | | |
| * | | | Add enough of a mechanism to allow us to update data in Mnesia but not the ↵Simon MacMullen2010-10-183-20/+33
| | | | | | | | | | | | | | | | | | | | schema.
| * | | | Detect failed upgrades.Simon MacMullen2010-10-181-7/+24
| | | | |
| * | | | Simplify the module attribute required, log a bit more.Simon MacMullen2010-10-183-33/+34
| | | | |
| * | | | DAG for upgrade steps.Simon MacMullen2010-10-154-15/+209
| | | | |
* | | | | Merged bug23299 (Erlang internode communication fails with user_sup undefined)Simon MacMullen2010-11-051-0/+5
|\ \ \ \ \
| * | | | | non_neg_integer(), surelySimon MacMullen2010-11-051-1/+1
| | | | | |
| * | | | | Add now_ms/0.Simon MacMullen2010-11-051-0/+5
| | | | | |
* | | | | | Make brokers launched from rabbitmq_multi.bat restartableEmile Joubert2010-11-051-4/+4
|/ / / / /
* | | | | bug23416 merged into defaultMarek Majkowski2010-11-0225-453/+586
|\ \ \ \ \
| * | | | | Changing opaque type to normal type as it at least stops dialyzer from ↵Matthew Sackman2010-11-021-6/+6
| | | | | | | | | | | | | | | | | | | | | | | | exploding. Dialyzer's findings are still wrong, but still...
| * | | | | Merge bug23454 into default (Enabling stats prevents hibernation, resulting ↵Simon MacMullen2010-11-023-5/+6
| |\ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | in excessive memory use)
| | * | | | | Channel was also wrong.Marek Majkowski2010-11-021-3/+4
| | | | | | |
| | * | | | | amqqueue emit_stats should reply with 'hibernate'Marek Majkowski2010-11-012-2/+2
| | | | | | |
View Lorry Controller Status