From c21f5d3b28f15a9872453f6968fdc91688f130ed Mon Sep 17 00:00:00 2001
From: Michal Arbet <michal.arbet@ultimum.io>
Date: Mon, 23 Dec 2024 01:44:43 +0100
Subject: [PATCH] Add retry mechanism to Ceph OSD addition task

This change ensures the `Add OSDs individually`
task is retried up to 3 times with a 10-second delay
between attempts if the Ceph orchestrator command fails
(non-zero return code). This enhances task resilience
by allowing transient issues to resolve before marking
the operation as failed.

Resolves stackhpc/ansible-collection-cephadm#161
---
 roles/cephadm/tasks/osds.yml | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/roles/cephadm/tasks/osds.yml b/roles/cephadm/tasks/osds.yml
index bb69137..925aaeb 100644
--- a/roles/cephadm/tasks/osds.yml
+++ b/roles/cephadm/tasks/osds.yml
@@ -20,3 +20,6 @@
     # NOTE: Without this, the delegate hosts's ansible_host variable will not
     # be respected.
     ansible_host: "{{ mon_ansible_host if 'mons' in group_names else hostvars[groups['mons'][0]].ansible_host }}"
+  until: osd_add_result.rc == 0
+  retries: 3
+  delay: 10