Ceph Administration – Master Open Source Distributed Storage
Formation en ligne
3 days (21 hours)

Ceph Administration – Master Open Source Distributed Storage

Learn to deploy and manage Ceph storage clusters to build reliable, high-performance, and scalable distributed storage infrastructure. Hands-on 3-day training covering RADOS architecture, core components, and basic administration.

Objectifs de la formation

Upon completion of this training, you will be able to:

  • Understand Ceph architecture and the RADOS (Reliable Autonomic Distributed Object Store) system
  • Install and configure a functional Ceph cluster using Cephadm
  • Manage core components: Monitors, OSDs, Managers, and MDS
  • Implement block storage with RADOS Block Device (RBD)
  • Deploy object storage via RADOS Gateway (RGW) with S3 compatibility
  • Configure file storage using CephFS
  • Administer pools and manage data replication
  • Use the Ceph dashboard for graphical monitoring
  • Diagnose and resolve common issues
  • Apply best practices for performance and security

Public concerné

This training is designed for:

System Administrators

Linux professionals looking to evolve into distributed storage administration and acquire Software-Defined Storage skills

Infrastructure Engineers

Responsible for deploying and managing scalable storage infrastructures for modern datacenters

Cloud Architects

Experts seeking a high-performance open source storage solution as an alternative to expensive proprietary solutions

OpenStack/Kubernetes Administrators

Professionals wanting to integrate Ceph as a storage backend for their cloud platforms

DevOps Engineers

Developers and operations staff looking to automate storage provisioning with Ansible

Training adapted for French-speaking African professionals seeking sovereign and cost-effective storage solutions.

Prérequis

Technical Prerequisites

Required

  • Linux Administration: Proficiency in basic commands, systemd service management, file editing
  • Storage Concepts: Understanding of basic concepts (partitions, file systems, RAID)
  • TCP/IP Networking: Fundamental knowledge (IP addressing, ports, protocols)
  • Shell Scripting: Ability to read and understand simple Bash scripts

Required Hardware Configuration

For online hands-on labs, each participant must have:

  • Stable internet connection (minimum 10 Mbps)
  • Modern web browser (Chrome, Firefox, Edge)
  • SSH client (Terminal, PuTTY, or equivalent)
  • Lab environment provided in ECINTELLIGENCE cloud

Programme détaillé

Detailed Training Program

Day 1: Fundamentals and Ceph Architecture

Module 1: Introduction to Distributed Storage and Ceph (3h)

  • Storage evolution: from DAS/NAS/SAN to Software-Defined Storage
  • Traditional storage challenges and distributed storage advantages
  • Ceph project history and philosophy
  • RADOS architecture: the heart of Ceph
  • Key concepts: pools, placement groups (PGs), CRUSH map
  • Use cases and success stories in Africa
  • Comparison with other solutions (GlusterFS, MinIO)
Hands-on Lab:

Exploring a demo Ceph cluster, analyzing the architecture

Module 2: Essential Ceph Components (4h)

  • Monitors (MON): Cluster map and quorum management
    • Role in cluster consistency
    • Quorum configuration (3, 5, 7 monitors)
    • Epoch and version management
  • OSDs (Object Storage Daemons): Data storage
    • Disk and partition management
    • Replication and recovery process
    • Journal and metadata
  • Managers (MGR): Metrics and plugins
    • Web dashboard
    • Prometheus, Zabbix modules
    • RESTful API
  • MDS (Metadata Server): For CephFS
    • Metadata cache
    • Active/Standby configuration
Hands-on Lab:

Installing a 3-node Ceph cluster with Cephadm (Reef 18.2)

Day 2: Deployment and Storage Types

Module 3: Deployment with Cephadm (3h)

  • System and network prerequisites for Ceph
  • Cluster bootstrap with cephadm
  • Adding nodes and OSDs
  • Configuring public and cluster networks
  • Service management with orchestrator
  • Upgrading to Squid 19.2
  • Deployment strategies: All-in-one vs distributed
Hands-on Lab:

Complete cluster deployment, adding nodes, network configuration

Module 4: Block Storage with RBD (2h)

  • RBD architecture and operation
  • Creating and managing RBD images
  • Snapshots and clones
  • Advanced features: layering, exclusive-lock
  • Integration with hypervisors (KVM/QEMU)
  • Mapping volumes on Linux clients
  • Performance tuning for database workloads
Hands-on Lab:

Creating RBD images, snapshots, client mounting, performance testing

Module 5: Object Storage with RGW (2h)

  • Object storage concepts: buckets and objects
  • RADOS Gateway deployment
  • Multi-site and zone configuration
  • S3 and Swift APIs: compatibility and differences
  • User and quota management
  • Bucket policies and ACLs
  • Use cases: archiving, backup, CDN
Hands-on Lab:

RGW deployment, bucket creation, testing with AWS CLI and s3cmd

Day 3: Advanced Administration and Optimization

Module 6: File Storage with CephFS (2h)

  • CephFS architecture: MDS and data/metadata pools
  • MDS deployment and configuration
  • Mounting CephFS: kernel client vs FUSE
  • Subvolume and snapshot management
  • Quotas and layouts
  • NFS and SMB export via Ganesha
  • Use cases: file sharing, home directories
Hands-on Lab:

CephFS configuration, client mounting, load testing

Module 7: Administration and Monitoring (3h)

  • Pool and CRUSH Management
    • Pool types: replicated vs erasure coded
    • CRUSH map modification
    • Custom placement rules
  • Dashboard and Monitoring
    • Navigating the Ceph dashboard
    • Key metrics: IOPS, throughput, latency
    • Prometheus/Grafana integration
  • Daily Maintenance
    • Essential commands: ceph status, health, df
    • OSD management: in/out, up/down
    • Scrubbing and deep-scrub processes
Hands-on Lab:

Monitoring configuration, creating custom dashboards

Module 8: Performance and Troubleshooting (2h)

  • Performance analysis with ceph bench
  • Optimization: cache tiering, optimal placement
  • Troubleshooting: slow requests, PG states
  • Recovery and backfilling: management and priorities
  • Practical failure scenarios and resolutions
  • Backup and disaster recovery strategies
  • Future developments: NVMe-oF, compression
Final Project:

Failure simulation and recovery, production cluster optimization

Certification and Assessment

  • Knowledge validation quiz at the end of training
  • ECINTELLIGENCE training certificate
  • Complete course materials in English (250+ pages)
  • Lab access for 30 days after training
  • Optional preparation for Red Hat Ceph certification
  • Integration into the French-speaking Africa Ceph community

Certification

À l'issue de cette formation, vous recevrez une attestation de participation délivrée par ECINTELLIGENCE.

1680 EUR

par participant

Durée

3 days (21 hours)

Modalité

Formation en ligne

Prochaine session

Sur demande

Demander un devis

Autres formations qui pourraient vous intéresser

Prêt à développer vos compétences ?

Rejoignez des centaines de professionnels qui ont fait confiance à ECINTELLIGENCE pour leur montée en compétences.

Voir toutes nos formations

Nathan

Assistant virtuel ECINTELLIGENCE