summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Have a completely separate counter for content size vs structural size.Simon MacMullen2014-03-245-32/+47
|
* Fix bitstrings, export term/3.Simon MacMullen2014-03-241-3/+7
|
* tiny tidyingSimon MacMullen2014-03-241-4/+4
|
* More robustness in the shape of unusual events.Simon MacMullen2014-03-241-6/+7
|
* Those aren't log eventsSimon MacMullen2014-03-241-2/+1
|
* cosmeticSimon MacMullen2014-03-241-1/+1
|
* Several more changes:Simon MacMullen2014-03-241-20/+19
| | | | | | * Remove the special tuple handling, it was trying to not reduce depth as we went "along" a tuple, but that goes badly in real world tests with gb_trees and gb_sets. In fact we just want to treat a tuple exactly like a list. * Simplify and correctify a bit by only having one termination condition for depth. * Add some more tests, and correct some existing ones.
* Something of a rewrite:Simon MacMullen2014-03-246-50/+79
| | | | | | | * Allow choice of how much size to "eat" each time we descend * Fix at least one more recurse-indefinitely bug * Try to reduce the amount of copying we do, and use more efficient techniques * Add some tests
* Remove obsolete TODOs, fix a rather substantial bug.Simon MacMullen2014-03-241-3/+1
|
* Don't let the truncate module have any opinion on how much to truncate.Simon MacMullen2014-03-245-16/+24
|
* Rename module and functions to be shorter and more generic. Remove ↵Simon MacMullen2014-03-245-183/+35
| | | | quickcheck tests. Wire into the non-sasl error_handler, but don't wire into handle_info and handle_call functions since those should not be logs.
* Tweak a bit more - attempt to handle bitstrings properlyTim Watson2014-03-212-8/+11
|
* Tweak list handlingTim Watson2014-03-212-21/+23
|
* Avoid crashing as we approach Len == 0, improve test outputTim Watson2014-03-212-5/+10
|
* Reduce the size of the test inputsTim Watson2014-03-211-1/+1
| | | | Don't hit 110% CPU (thanks for that PropEr) and get usable failures sooner.
* Skip tests in a more operator friendly wayTim Watson2014-03-211-24/+23
|
* Start testing this properlyTim Watson2014-03-212-1/+147
|
* Avoid binary:part/3, since it isn't available in R13Tim Watson2014-03-211-2/+3
|
* Initial stab at limiting term size during loggingTim Watson2014-03-203-4/+86
|
* Merge bug25374Simon MacMullen2014-03-19151-247/+391
|\
| * Merge bug25722 (again)Simon MacMullen2014-03-191-1/+1
| |\
| | * Better error catching.Simon MacMullen2014-03-191-1/+1
| | |
| * | Merge bug25722 (again again).Simon MacMullen2014-03-18150-288/+312
| |\ \ | | |/ | |/|
| | * Merge bug25722 (again)Simon MacMullen2014-03-182-4/+5
| | |\
| | * \ Merge bug25722Simon MacMullen2014-03-1813-93/+231
| | |\ \
| | * \ \ Merge bug 25940Simon MacMullen2014-03-17150-152/+152
| | |\ \ \
| | | * | | Update copyright for 2014Simon MacMullen2014-03-17150-152/+152
| | |/ / /
| * | | | Don't explode on "rabbitmqctl stop_app ; rabbitmqctl status", and fix tests.Simon MacMullen2014-03-182-3/+8
| | |_|/ | |/| |
| * | | We want reuseaddr otherwise starting the same broker twice in quick ↵Simon MacMullen2014-03-181-1/+1
| | | | | | | | | | | | | | | | succession can fail for no good reason here - the real distribution mechanism obviously uses something similar.
| * | | Don't export anything ever from this script, it gets inherited by any ↵Simon MacMullen2014-03-181-3/+4
| | |/ | |/| | | | | | | processes started by the broker - which breaks a bunch of our tests.
| * | Oops, we need RABBITMQ_CONFIG_FILE to be set for prelaunch to do its job ↵Simon MacMullen2014-03-181-4/+4
| | | | | | | | | | | | properly.
| * | Merge in defaultSimon MacMullen2014-03-141-4/+8
| |\ \
| * | | Check the config situation when installing the Windows service.Simon MacMullen2014-03-132-11/+41
| | | |
| * | | Update the batch file.Simon MacMullen2014-03-131-1/+14
| | | |
| * | | More robustnessSimon MacMullen2014-03-131-16/+26
| | | |
| * | | Move dist arg to the end so that if the user has set it in ↵Simon MacMullen2014-03-131-1/+1
| | | | | | | | | | | | | | | | RABBITMQ_SERVER_START_ARGS then we take their version not ours.
| * | | Check if dist port is available since otherwise we will create a horrific ↵Simon MacMullen2014-03-132-20/+38
| | | | | | | | | | | | | | | | error message. Also refactor a touch.
| * | | CosmeticSimon MacMullen2014-03-131-7/+7
| | | |
| * | | Somehow that actually worked on OS X. Oops.Simon MacMullen2014-03-131-2/+2
| | | |
| * | | Try to be marginally smarter about getting the epmd port, it is set as an ↵Simon MacMullen2014-03-131-5/+7
| | | | | | | | | | | | | | | | init argument from the environment variable.
| * | | Present epmd errors a bit better.Simon MacMullen2014-03-121-7/+12
| | | |
| * | | Display diagnostics when rabbitmqctl clustering options fail due to nodedown.Simon MacMullen2014-03-122-37/+23
| | | |
| * | | That should probably be a define, right?Simon MacMullen2014-03-121-1/+3
| | | |
| * | | Irritatingly I can't find a way to figure out the difference between these ↵Simon MacMullen2014-03-121-0/+1
| | | | | | | | | | | | | | | | programmatically.
| * | | Slightly cleaner than iterating through families.Simon MacMullen2014-03-121-8/+10
| | | |
| * | | Don't suggest using the kernel params here.Simon MacMullen2014-03-121-9/+1
| | | |
| * | | Set the distribution port range to the first of these that matches:Simon MacMullen2014-03-122-10/+39
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Not at all if it's mentioned in the config file * $DIST_PORT if that's mentioned in the env file * $RABBITMQ_DIST_PORT if that is set * $RABBITMQ_NODE_PORT + 20000 if that is set * 25672 in all other cases
| * | | Just in case the error is not an atom.Simon MacMullen2014-03-121-1/+1
| | | |
| * | | Various formatting tweaks.Simon MacMullen2014-03-121-5/+4
| | | |
| * | | Since we always show the raw error and the expanded error let's get ↵Simon MacMullen2014-03-125-13/+14
| | | | | | | | | | | | | | | | format_inet_error to do that.
View Lorry Controller Status