Our documentation contain suboptimal update step sequence:
specific ceph packages are updated after all installed packages
were updated on slave nodes.
Another issue is late ceph restart that could lead to a broken
ceph cluster if customer will accidently restart node to solve
other update issues.
I have moved update ceph step earlier and changed the command to
restart ceph services right after package update.
Reviewed: https:/ /review. fuel-infra. org/35628
Submitter: Olena Logvinova <email address hidden>
Branch: stable/9.2
Commit: 8be814a372e1fc5 89cda57ebba3fa7 25b2cd0d99
Author: Alexey Stupnikov <email address hidden>
Date: Wed Jun 14 17:28:19 2017
Improve 9.2 update process
Our documentation contain suboptimal update step sequence:
specific ceph packages are updated after all installed packages
were updated on slave nodes.
Another issue is late ceph restart that could lead to a broken
ceph cluster if customer will accidently restart node to solve
other update issues.
I have moved update ceph step earlier and changed the command to
restart ceph services right after package update.
Change-Id: Ib44e10256c41a2 b182b80c8a3d870 9af8454c47f
Closes-bug: #1661260