From adfb56a8faafbce8547d0c3da2fa6921104607df Mon Sep 17 00:00:00 2001
From: Mark Goddard <mark@stackhpc.com>
Date: Thu, 8 Aug 2024 12:36:54 +0100
Subject: [PATCH] Add ceph-{enter,exit}-maintenance.yml playbooks

These can be used to enter and exit maintenance for Ceph hosts.
---
 doc/source/operations/upgrading-ceph.rst          |  7 ++++---
 etc/kayobe/ansible/ceph-enter-maintenance.yml     | 13 +++++++++++++
 etc/kayobe/ansible/ceph-exit-maintenance.yml      | 12 ++++++++++++
 etc/kayobe/ansible/requirements.yml               |  2 +-
 .../notes/ceph-maintenance-4c4eb0a4f7149665.yaml  | 15 +++++++++++++++
 5 files changed, 45 insertions(+), 4 deletions(-)
 create mode 100644 etc/kayobe/ansible/ceph-enter-maintenance.yml
 create mode 100644 etc/kayobe/ansible/ceph-exit-maintenance.yml
 create mode 100644 releasenotes/notes/ceph-maintenance-4c4eb0a4f7149665.yaml

diff --git a/doc/source/operations/upgrading-ceph.rst b/doc/source/operations/upgrading-ceph.rst
index 838a518a4..5ce711011 100644
--- a/doc/source/operations/upgrading-ceph.rst
+++ b/doc/source/operations/upgrading-ceph.rst
@@ -63,7 +63,7 @@ Place the host or batch of hosts into maintenance mode:
 
 .. code-block:: console
 
-   sudo cephadm shell -- ceph orch host maintenance enter <host>
+   kayobe playbook run $KAYOBE_CONFIG_PATH/ansible/ceph-enter-maintenance.yml -l <host>
 
 To update all eligible packages, use ``*``, escaping if necessary:
 
@@ -72,7 +72,8 @@ To update all eligible packages, use ``*``, escaping if necessary:
    kayobe overcloud host package update --packages "*" --limit <host>
 
 If the kernel has been upgraded, reboot the host or batch of hosts to pick up
-the change:
+the change. While running this playbook, consider setting ``ANSIBLE_SERIAL`` to
+the maximum number of hosts that can safely reboot concurrently.
 
 .. code-block:: console
 
@@ -82,7 +83,7 @@ Remove the host or batch of hosts from maintenance mode:
 
 .. code-block:: console
 
-   sudo cephadm shell -- ceph orch host maintenance exit <host>
+   kayobe playbook run $KAYOBE_CONFIG_PATH/ansible/ceph-exit-maintenance.yml -l <host>
 
 Wait for Ceph health to return to ``HEALTH_OK``:
 
diff --git a/etc/kayobe/ansible/ceph-enter-maintenance.yml b/etc/kayobe/ansible/ceph-enter-maintenance.yml
new file mode 100644
index 000000000..c119583e1
--- /dev/null
+++ b/etc/kayobe/ansible/ceph-enter-maintenance.yml
@@ -0,0 +1,13 @@
+---
+- name: Ensure a Ceph host has entered maintenance
+  gather_facts: true
+  any_errors_fatal: true
+  # We need to check whether it is OK to stop hosts after previous hosts have
+  # entered maintenance.
+  serial: 1
+  hosts: ceph
+  become: true
+  tasks:
+    - name: Ensure a Ceph host has entered maintenance
+      ansible.builtin.import_role:
+        name: stackhpc.cephadm.enter_maintenance
diff --git a/etc/kayobe/ansible/ceph-exit-maintenance.yml b/etc/kayobe/ansible/ceph-exit-maintenance.yml
new file mode 100644
index 000000000..a16f338d9
--- /dev/null
+++ b/etc/kayobe/ansible/ceph-exit-maintenance.yml
@@ -0,0 +1,12 @@
+---
+- name: Ensure a Ceph host has exited maintenance
+  gather_facts: true
+  any_errors_fatal: true
+  hosts: ceph
+  # The role currently requires hosts to exit maintenance serially.
+  serial: 1
+  become: true
+  tasks:
+    - name: Ensure a Ceph host has exited maintenance
+      ansible.builtin.import_role:
+        name: stackhpc.cephadm.exit_maintenance
diff --git a/etc/kayobe/ansible/requirements.yml b/etc/kayobe/ansible/requirements.yml
index 90e653ab2..2bb4047ce 100644
--- a/etc/kayobe/ansible/requirements.yml
+++ b/etc/kayobe/ansible/requirements.yml
@@ -1,7 +1,7 @@
 ---
 collections:
   - name: stackhpc.cephadm
-    version: 1.15.1
+    version: 1.18.0
   # NOTE: Pinning pulp.squeezer to 0.0.13 because 0.0.14+ depends on the
   # pulp_glue Python library being installed.
   - name: pulp.squeezer
diff --git a/releasenotes/notes/ceph-maintenance-4c4eb0a4f7149665.yaml b/releasenotes/notes/ceph-maintenance-4c4eb0a4f7149665.yaml
new file mode 100644
index 000000000..285e396f8
--- /dev/null
+++ b/releasenotes/notes/ceph-maintenance-4c4eb0a4f7149665.yaml
@@ -0,0 +1,15 @@
+---
+features:
+  - |
+    Adds two new custom playbooks for placing Ceph hosts into and removing them
+    from maintenance:
+
+    - ``ceph-enter-maintenance.yml``
+    - ``ceph-exit-maintenance.yml``
+upgrade:
+  - |
+    Updates the ``stackhpc.cephadm`` collection to version ``1.18.0``.
+fixes:
+  - |
+    Fixes an issue with idempotency in the ``stackhpc.ceph.cephadm_keys``
+    plugin.