Tips and Tricks for Managing and Administering Ceph Clusters
summary
Cephadm (Octopus) provisions a containerized Ceph cluster from bootstrap to working MON and manager in about two minutes and then adds hosts over SSH, with the first five hosts defaulting to monitors (configurable to three). It requires python3, LVM and a container runtime and can migrate clusters from Ceph Ansible, perform automated upgrades, and deploy OSDs with an activate all scan of devices. The Ceph Dashboard is installed by default and uses Prometheus and Grafana containers for metrics and alerts.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 0:48Background
- 1:25Prerequisites
- 2:20What is Ceph Manager
- 3:55Why Ceph Manager
- 7:25Bootstrapping
- 9:02Basic Overview
- 10:43Documentation
- 10:55Ceph Dashboard
- 11:35Ceph Dashboard Overview
- 12:47Ceph Dashboard Slides
related talks
29mTroubleshooting a Cephadm Deployment
05mHealthier Ceph Clusters with Ceph-medic
28mA Welcoming Ceph Experience Improvements for First-Time Users or: How to Make Beginners Love Ceph
31mMonitoring and Centralized Logging in Ceph
40mJuggling Petabytes: Managing Ceph at Scale with Ceph-ansible
36m200 Clusters vs 1 Admin
28mAutomatic Operation Bot for Ceph
44mContainer Management at Scale