Run ceph on two nodes, using one of them as a client,
with a separate client-only node.
Use xfs beneath the osds.
- install ceph/octopus v15.2.2 and the v15.2.x point versions
- install ceph/octopus v15.2.1 point version
++ install ceph/octopus v15.2.3 and the v15.2.x point versions
run workload and upgrade-sequence in parallel
(every point release should be tested)
run workload and upgrade-sequence in parallel
- print: "**** done workload v15.2.1"
- ####### upgrade to v15.2.2
-####### upgrade to v15.2.?? PLACEHOLDER
-#- install.upgrade:
-# #exclude_packages: ['ceph-mgr','libcephfs2','libcephfs-devel','libcephfs-dev']
-# mon.a:
-# tag: v15.2.??
-# mon.b:
-# tag: v15.2.??
-# # Note that client.a IS NOT upgraded at this point
-#- parallel:
-# - workload_octopus
-# - upgrade-sequence_octopus
-#- print: "**** done parallel octopus v15.2.??"
++####### upgrade to v15.2.3
+- install.upgrade:
+ #exclude_packages: ['ceph-mgr','libcephfs2','libcephfs-devel','libcephfs-dev']
+ mon.a:
- tag: v15.2.2
++ tag: v15.2.3
+ mon.b:
- tag: v15.2.2
++ tag: v15.2.3
+- parallel:
+ - workload_octopus
+ - upgrade-sequence_octopus
- - print: "**** done parallel octopus v15.2.2"
++- print: "**** done parallel octopus v15.2.3"
#### upgrade to latest octopus
- install.upgrade: