summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Rename aliveness_test to node_health_checkDiana Corbacho2016-02-252-24/+24
|
* Use net_ticktime + 10s as timeout in the aliveness test and abort with the ↵Diana Corbacho2016-02-251-14/+15
| | | | first error
* Add aliveness_test documentationDiana Corbacho2016-02-251-0/+15
|
* Aliveness checkDiana Corbacho2016-02-251-1/+64
| | | | * Basic checks
* Merge pull request #631 from rabbitmq/rabbitmq-server-625Michael Klishin2016-02-171-0/+2
|\ | | | | Added erl call to make sure epmd is started
| * IndentationMichael Klishin2016-02-171-2/+2
| |
| * added erl with -sname to make sure epmd is startedgabriele santomaggio2016-02-171-0/+2
|/
* Use basic regular expressions for sed portabilityChris Bandy2016-02-152-12/+3
|
* Merge pull request #617 from rabbitmq/rabbitmq-server-616Michael Klishin2016-02-111-3/+56
|\ | | | | rabbit_mirror_queue_sync: Pause the syncer process if memory is low
| * rabbit_mirror_queue_sync: Pause the syncer process if memory is lowJean-Sébastien Pédron2016-02-101-3/+56
|/ | | | | | | If any node in the cluster is running out of memory, we pause the syncer process. If we don't do this, this alarmed node could crash. Fixes #616.
* rabbit_alarm: Fix resource_alarm_source() typeJean-Sébastien Pédron2016-02-101-1/+1
| | | | | Alarms are emitted for 'disk' or 'memory' resources shortage, not 'node'.
* rabbitmq-env.bat: Do not prepare and escape RABBITMQ_PLUGINS_EXPAND_DIRrabbitmq_v3_6_1_rc1Jean-Sébastien Pédron2016-02-091-4/+4
| | | | | | | | RabbitMQ recreates it anyway so we can't pass it the escaped directory name. Therefore, this directory must not contain any non-US-ASCII in its name. References #493.
* Merge pull request #609 from bogdando/bug/1541029Michael Klishin2016-02-051-15/+25
|\ | | | | Fix action_stop for the rabbit OCF
| * Fix action_stop for the rabbit OCFBogdan Dobrelya2016-02-041-15/+25
|/ | | | | | | | | | | | | The action_stop may sometimes stop the rabbitmq-server gracefully by the PID, but leave unresponsive beam.smp processes running and spoiling rabbits. Those shall be stopped as well. The solution is: - make proc_stop() to accept a pid=none to use a name matching instead - make kill_rmq_and_remove_pid() to stop by the beam process matching as well - fix stop_server_process() to ensure there is no beam process left running Related Fuel bug: https://launchpad.net/bugs/1541029 Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>
* Merge pull request #604 from rabbitmq/rabbitmq-server-595Michael Klishin2016-02-041-1/+3
|\ | | | | Add emitting node info to amq.rabbitmq.log messages
| * Add emitting node info to amq.rabbitmq.log messagesMichael Klishin2016-02-041-1/+3
|/ | | | Fixes #595.
* Merge pull request #602 from dmitrymex/stableMichael Klishin2016-02-031-23/+21
|\ | | | | Streamline checking for cluster partitioning
| * Streamline checking for cluster partitioningDmitry Mescheryakov2016-02-031-23/+21
|/ | | | | Move check if we are current cluster master to earlier place in code. That way we will avoid unnecessary operations for master case.
* Merge pull request #601 from bogdando/bug/1540936Michael Klishin2016-02-031-0/+5
|\ | | | | Fix cluster membership check for running master
| * Fix cluster membership check for running masterBogdan Dobrelya2016-02-021-0/+5
| | | | | | | | The running master is always inside of its own cluster. Fix the cluster membership check when a node is the master.
* | Merge pull request #600 from bogdando/bug/1540936Michael Klishin2016-02-021-0/+1
|\ \ | |/ | | Fix uninitialized status_master
| * Fix uninitialized status_masterBogdan Dobrelya2016-02-021-0/+1
|/ | | | | | | | Fix multiple nodes may be reported in logs as the running master Related Fuel bug https://bugs.launchpad.net/bugs/1540936 Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>
* Merge pull request #597 from binarin/rabbitmq-server-ocf-quiet-curlMichael Klishin2016-02-021-1/+1
|\ | | | | Suppress curl progress indicator in rabbit OCF
| * Suppress curl progress indicator in rabbit OCFAlexey Lebedeff2016-02-021-1/+1
|/ | | | | | | | | | | | | curl is used by OCF script for fetching definitions (queues etc.), but results of that invocation is shown as garbage in pacemaker logs - progress indicator doesn't make any sense in logs. According to curl manpage the following combination of options should be used "--silent --show-error" - this will suppress only progress indicator, errors will still be shown. Also other short curl options are replaced with their long counterparts - for improved readability.
* Merge pull request #594 from rabbitmq/rabbitmq-server-592Michael Klishin2016-02-011-1/+6
|\ | | | | Use -r with sed on Linux
| * Use -r with sed on Linux, fixes #592Michael Klishin2016-02-011-1/+6
|/ | | | | | | | We previously did the same change in #273 (PR: #275), but the file in which it was done was removed in 231e90cacf3daec5f43b3307867129e61496b123. Note that #592 recommends using `-r` unconditionally but that option is not recognised by sed which ships with OS X.
* Merge branch 'rabbitmq-server-541' into stableMichael Klishin2016-01-291-4/+6
|\ | | | | | | | | | | | | | | This is only a part of what #541 is supposed to cover but it already helped in a particular node shutdown lockup we've observed => worth merging earlier. Per discussion with @dcorbacho.
| * Merge branch 'stable' into rabbitmq-server-541Michael Klishin2016-01-271-1/+9
| |\
| * | Increase supervisor timeout to net_ticktime + 10sDiana Corbacho2016-01-271-1/+1
| | |
| * | Introduce timeout in rabbit_channel_supDiana Corbacho2016-01-271-4/+6
| | |
* | | Merge pull request #588 from rabbitmq/rabbitmq-management-117Michael Klishin2016-01-291-4/+12
|\ \ \ | |_|/ |/| | Specify hash algorithm in change_password_hash
| * | Trailing wsMichael Klishin2016-01-281-3/+3
| | |
| * | Specify hash algorithm in change_password_hashDaniil Fedotov2016-01-281-3/+11
|/ /
* | Merge pull request #584 from rabbitmq/rabbitmq-server-581Michael Klishin2016-01-271-1/+9
|\ \ | |/ |/| Unblock receive after 15s
| * Unblock receive after 15sDiana Corbacho2016-01-271-1/+9
| |
* | rabbit.erl: Do not run systemd-notify on WindowsJean-Sébastien Pédron2016-01-271-4/+7
| | | | | | | | This silences a warning logged during RabbitMQ startup.
* | Merge pull request #535 from rabbitmq/rabbitmq-server-307Jean-Sébastien Pédron2016-01-271-0/+1
|\ \ | | | | | | Ignore duplicate down_from_ch
| * | Ignore duplicate down_from_chDaniil Fedotov2016-01-081-0/+1
| | |
* | | Merge branch 'rabbitmq-server-493' into stableMichael Klishin2016-01-276-24/+121
|\ \ \ | |_|/ |/| |
| * | Create directories and files on Windows before conversion to short filenamesJean-Sébastien Pédron2016-01-262-2/+48
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If the directory or file does not exist before RabbitMQ starts, we can't let RabbitMQ create it, otherwise, it's created with its short filename, not its long one. With this new correction, we can "escape" all variables instead of only RABBITMQ_BASE. Fixes #493.
| * | rabbitmq-server.bat: Honor RABBITMQ_LOGS=- to log to stdoutJean-Sébastien Pédron2016-01-261-3/+21
| | | | | | | | | | | | | | | | | | | | | | | | Note that at the time of this commit, Lager does not support logging to stdout on Windows. This commit still improves consistency between Unix and Windows. References #493.
| * | Use RABBITMQ_HOME to set the path to RabbitMQ ebin directoryJean-Sébastien Pédron2016-01-264-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | Compared to the script's parent directory (stored in TDP0), RABBITMQ_HOME is converted to a short filename to avoid non-ASCII in the path. Fixes #493.
| * | Use short filenames in Windows startup scriptsJean-Sébastien Pédron2016-01-262-17/+50
|/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On Windows, cmd.exe and batch scripts do not support Uniode apparently. However, Windows uses UTF-16 to encode filenames one disk. In batch scripts, filenames are converted to some one-byte-wide charset. Once passed to Erlang and RabbitMQ, those filenames are incorrect. In particular, the management UI is unhappy because filenames obviously contain invalid UTF-8 characters. Using short filenames makes sure filename only contain US-ASCII characters. To convert them, we use "for" expansion. At the same time, filenames are made absolute. It works even better than realpath.exe because the latter also converts filenames to another charset again. Fixe #493.
* | Merge pull request #573 from binarin/rabbitmq-server-systemd-notify-zero-depsMichael Klishin2016-01-221-1/+1
|\ \ | | | | | | Use systemd-notify(1) shell helper as fallback
| * | Use systemd-notify(1) shell helper as fallbackAlexey Lebedeff2016-01-221-1/+1
|/ / | | | | | | | | | | | | | | | | | | | | | | Currently external erlang library `sd_notify` is used to make systemd unit with `Type=notify` to work correctly. This library contains some C code and thus cannot be built into architecture-independent package. But it is not actually needed, as systemd provides systemd-notify(1) helper for shell scripts which serves exactly the same purpose. The only thing is that you need to add `NotifyAccess=all` to your unit file to make everything work well.
* | Merge pull request #571 from binarin/rabbitmq-server-ocf-shell-quotingMichael Klishin2016-01-211-1/+1
|\ \ | | | | | | Fix usage of uninitialized variable in OCF script
| * | Fix usage of uninitialized variable in OCF scriptAlexey Lebedeff2016-01-211-1/+1
|/ /
* | Merge pull request #563 from ↵Michael Klishin2016-01-211-0/+109
|\ \ | | | | | | | | | | | | binarin/rabbitmq-server-ocf-list-channels-diagnostics Improve OCF script diagnostics for timed-out 'list_channels'
| * | Improve 'list_channels' diagnostics in OCFAlexey Lebedeff2016-01-201-1/+1
| | | | | | | | | timeout(1) manpage mentions 124 as another valid return code from, in addition to 128 + signal-number.
| * | Improve rabbitmq OCF script diagnosticsAlexey Lebedeff2016-01-201-0/+109
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently time-out when running 'rabbitmqctl list_channels' is treated as a sign that current node is unhealthy. But it could not be the case, as the hanging channel could be actually on some other node. Given that currently we have more than one bug related to 'list_channels', it makes sense to improve diagnostics here. This patch doesn't change any behaviour, only improves logging after time-out happens. If time-outs continue to occur (even with latest rabbitmq versions or with backported fixes), we could switch to this improved list_channels and kill rabbitmq only if stuck channels are located on current node. But I hope that all related rabbitmq bugs were already closed.
View Lorry Controller Status