Fix typo when nova pool is built and wrong flag during cli creation
When the upgrade from nautilus to pacific is performed, the regular cephadm playbook fails during pools creation. This is caused due to a typo in the build_pools task that processes the vms pool. In addition, during the same run, if no spec is present, post.yaml fails getting the right cli due to a wrong 'mount_spec' flag value. This change fixes both failures removing the extra space that caused the pool creation failure and the mount_spec flag that caused an issue in the get_cli task. Closes-Bug: #1990195 Signed-off-by: Francesco Pantano <fpantano@redhat.com> Change-Id: I253c8feb72404e2620821472447b08905d5cd16b
This commit is contained in:
parent
948a5f23d8
commit
b200d84dd1
|
@ -16,8 +16,6 @@
|
|||
|
||||
- name: Get ceph_cli
|
||||
include_tasks: ceph_cli.yaml
|
||||
vars:
|
||||
mount_spec: true
|
||||
|
||||
- name: Get the ceph orchestrator status
|
||||
command: "{{ tripleo_cephadm_ceph_cli }} orch status --format json"
|
||||
|
|
|
@ -37,7 +37,7 @@
|
|||
|
||||
- name: add vms pool
|
||||
set_fact:
|
||||
vms: [{'name': '{{ ceph_pools.nova_pool.name }}', 'pg_num': '{{ ceph_pools.pg_num | default(omit) }} ', 'rule_name': 'replicated_rule', 'application': 'rbd'}]
|
||||
vms: [{'name': '{{ ceph_pools.nova_pool.name }}', 'pg_num': '{{ ceph_pools.pg_num | default(omit) }}', 'rule_name': 'replicated_rule', 'application': 'rbd'}]
|
||||
when:
|
||||
- tripleo_pool_vms
|
||||
- ceph_pools.nova_pool.name not in overrides|default([])
|
||||
|
|
Loading…
Reference in New Issue