Upgrade Ceph Cluster

Upgrade Ceph Cluster

Introduction

Upgrading your Ceph cluster is a crucial process to ensure you are leveraging the latest features, security updates, and performance improvements. This document provides a step-by-step guide to upgrading your Ceph cluster, focusing on transitioning to version 18.2.2. The process involves verifying cluster health, initiating the upgrade, and monitoring the progress to ensure a smooth and successful upgrade. By following these procedures, you can maintain the robustness and efficiency of your Ceph storage solution.

Procedure 1

  1. Ensure your cluster is healthy and all nodes are online

    bash
    ceph -s
  2. Run ceph orch upgrade start command 2

bash
ceph orch upgrade start --ceph-version <version>

Example: upgrade to version 18.2.2

bash
ceph orch upgrade start --ceph-version 18.2.2

Monitor Upgrade

You can use the ceph orch upgrade status command to check if the upgrade is in progress. You can also utlize the ceph -s (ceph -w) command to watch the progress bar and receive updates as to the progress of the upgrade.

bash
ceph -w

Output

bash
cluster:
  id: ad99432c-437e-11ef-8f76-86577d660b4a
  health: HEALTH_OK

services:
  mon: 3 daemons, quorum host01, host02, host03 (age 5w)
  mgr: host01.fhdyeg(active, since 13d), standbys: host2.vhndye, host03.avdefo
  osd: 36 osds: 36 up (since 4w), 36 in (since 4w)

data:
  pools:   6 pools, 161 pgs
  objects: 8.19k objects, 32 GiB
  usage:   18 TiB used, 655 TiB / 672 TiB avail
  pgs :    161 active+clean

progress:
  Upgrade to 18.2.2 (21s)
    [===.........................]

 2024-07-16T10:16:44.270101-0400 mon. host01 [INF] Manager daemon host01.fhdyeg is now available
 2024-07-16T10:18:34.067874-0400 mon. host01 [INF] Manager daemon host2.vhndye is now available
 2024-07-16T10:19:41.460139-0400 mon. host01 [INF] mon. host01 calling monitor election
Last updated on