From 570c7ffc0149b9fbd20b50b54632f37990b27128 Mon Sep 17 00:00:00 2001 From: Patrick Donnelly Date: Wed, 21 Apr 2021 11:09:53 -0700 Subject: [PATCH] doc: update mds_autoscaler documentation to reflect recent changes Notably, Sage changed how this worked in [1]. [1] 6d8c0e4722eb5f3014dc76f6733afea08bcc8e72 Fixes: https://tracker.ceph.com/issues/51187 Signed-off-by: Patrick Donnelly --- doc/mgr/mds_autoscaler.rst | 36 ++++++++++++++---------------------- 1 file changed, 14 insertions(+), 22 deletions(-) diff --git a/doc/mgr/mds_autoscaler.rst b/doc/mgr/mds_autoscaler.rst index 39dadd882929..46fc44155224 100644 --- a/doc/mgr/mds_autoscaler.rst +++ b/doc/mgr/mds_autoscaler.rst @@ -1,31 +1,23 @@ MDS Autoscaler Module ===================== -The MDS Autoscaler Module monitors ``fsmap`` update notifications from the mgr -daemon and takes action to spawn or kill MDS daemons for a file-system as per -changes to the: +The MDS Autoscaler Module monitors file systems to ensure sufficient MDS +daemons are available. It works by adjusting the placement specification for +the orchestrator backend of the MDS service. To enable, use: -- ``max_mds`` config value -- ``standby_count_wanted`` config value -- standby promotions to active MDS state in case of active MDS rank death +.. sh: -Bumping up the ``max_mds`` config option value causes a standby mds to be promoted -to hold an active rank. This leads to a drop in standby mds count. The MDS -Autoscaler module detects this deficit and the orchestrator module is notified -about the required MDS count. The orchestrator back-end then takes necessary -measures to spawn standby MDSs. + ceph mgr module enable mds_autoscaler -Dropping the ``max_mds`` config option causes the orchestrator back-end to kill -standby mds to achieve the new reduced count. Preferably standby mds are chosen -to be killed when the ``max_mds`` count is dropped. +The module will monitor the following file system settings to inform +placement count adjustments: -An increment and decrement of the ``standby_count_wanted`` config option value -has a similar effect on the total MDS count. The orchestrator is notified about -the change and necessary action to spawn or kill standby MDSs is taken. +- ``max_mds`` file system setting +- ``standby_count_wanted`` file system setting -A death of an active MDS rank also causes promotion of a standby mds to occupy -the required active rank. The MDS Autoscaler notices the change in the standby -mds count and a message is passed to the orchestrator to maintain the necessary -MDS count. +The Ceph monitor daemons are still responsible for promoting or stopping MDS +according to these settings. The ``mds_autoscaler`` simply adjusts the +number of MDS which are spawned by the orchestrator. -NOTE: There is no CLI associated with the MDS Autoscaler Module. +.. note: There is no CLI or module configurations as of now. Enable or disable + the module to turn on or off. -- 2.47.3