3 Setup 4 node ceph cluster using ceph-deploy, use latest
4 stable jewel as initial release, upgrade to luminous and
5 also setup mgr nodes along after upgrade, check for
6 cluster to reach healthy state, After upgrade run kernel tar/untar
7 task and systemd task. This test will detect any
8 ceph upgrade issue and systemd issues.
14 mon pg warn min per osd: 2
16 osd pool default size: 2
17 osd objectstore: filestore
20 rbd default features: 5
27 # reluctantely :( hard-coded machine type
28 # it will override command line args with teuthology-suite
49 - print: "**** done ssh-keys"
54 - print: "**** done initial ceph-deploy"
55 - ceph-deploy.upgrade:
59 check-for-healthy: True
65 - print: "**** done ceph-deploy upgrade"
68 - ceph osd require-osd-release luminous
69 - ceph osd set-require-min-compat-client luminous
70 - print: "**** done `ceph osd require-osd-release luminous`"
74 - kernel_untar_build.sh
75 - print: "**** done kernel_untar_build.sh"
77 - print: "**** done systemd"
81 - rados/load-gen-mix.sh
82 - print: "**** done rados/load-gen-mix.sh"