How luisaostuff/san-basada-en-arm-y-alta-disponibilidad works
Project Type and Purpose
• This is an infrastructure‐as‐code project that defines and deploys a highly available SAN (Storage Area Network) cluster on three Odroid-hc1 (ARM) boards.
• It uses Ansible to automate the installation and configuration of DRBD (block‐device replication), Pacemaker (cluster resource manager), and iSCSI (exporting block devices to clients).
File Structure Analysis
• Top level:
• Markdown docs (design rationale, installation manuals, conclusions).
• Ansible directory (ansible.cfg, hosts inventory, roles folder, site.yml).
• recursos/scripts (benchmarking scripts).
• ansible/roles:
• commons – common host setup (hosts file, basic packages).
• drbd – DRBD resource definition, synchronization scripts.
• pacemaker – cluster configuration scripts, resource start/stop.
• site.yml orchestrates the run‐order of these roles across the three target nodes.
README Insights
• Three‐node ARM cluster for low‐power, scalable, HA storage.
• Ubuntu 20.04 LTS on each Odroid-hc1.
• DRBD for real‐time block replication.
• Pacemaker for failover and resource orchestration.
• iSCSI to expose LUNs to external clients.
• Optional future: advanced filesystem, monitoring stack, extended Ansible recipes.
a. Main Components
• Ansible Control Host (could be your laptop or a dedicated management VM)
• Three ARM Storage Nodes (Node-A, Node-B, Node-C)
• Network Segments:
• Management Network (SSH/Ansible control)
• DRBD Replication Network (private, for block‐device sync)
• Client Network (iSCSI‐client traffic)
• DRBD Block Devices (one on each node, forming a replicated resource)
• Pacemaker Cluster Service (running on each node, monitoring DRBD, managing failover)
• iSCSI Target Service (running on active node(s), exporting LUNs)
• Client Machines (ESXi, Linux, etc., consuming the iSCSI LUN)
b. Relationships and Interactions
• Ansible → (SSH) → All Nodes: pushes “commons” → DRBD → Pacemaker roles.
• Nodes ↔ Nodes (Replication Network): DRBD streaming of data between block devices.
• Nodes ↔ Nodes (Heartbeat/Corosync if used internally): Pacemaker cluster communication.
• Active Node → Clients: iSCSI connections for storage I/O.
c. Architectural Patterns & Design Principles
• Infrastructure as Code: roles for repeatable, idempotent configuration.
• High Availability Cluster: DRBD provides data redundancy; Pacemaker handles automatic failover.
• Separation of Planes:
• Control Plane – Ansible orchestration
• Data Plane – DRBD replication & iSCSI traffic
• Management Plane – Pacemaker/Corosync cluster heartbeats
d. Key Technologies
• Ansible (playbooks, roles)
• DRBD (Linux kernel module + userland)
• Pacemaker (cluster resource manager)
• iSCSI Target (e.g., tgt or LIO)
• Ubuntu 20.04 LTS on Odroid-hc1 (ARM architecture)