]> git.apps.os.sepia.ceph.com Git - ceph.git/commit
mgr/cephadm/upgrade: scale down MDS cluster(s) for major version upgrades
authorSage Weil <sage@newdream.net>
Tue, 26 Jan 2021 22:10:13 +0000 (16:10 -0600)
committerSage Weil <sage@newdream.net>
Mon, 1 Feb 2021 23:20:14 +0000 (17:20 -0600)
commit604d2636d7744b742c162df722ed53ec68a9084e
tree88bb063ecb72b2d641007e46f41169d9aacd5fd5
parent1e60a103709aec5ddac8b12a44b9d0d646d2a682
mgr/cephadm/upgrade: scale down MDS cluster(s) for major version upgrades

For octopus -> pacific, as with other recent releases, we need to scale
down the MDS cluster(s) to a single daemon before upgrading.  (This is
because the MDS intra-cluster protocols aren't fully versioned.)

Signed-off-by: Sage Weil <sage@newdream.net>
src/pybind/mgr/cephadm/upgrade.py