Skip to content

🧱 GlusterFS Multi-Node Cluster Storage

Gubernator provides enterprise distributed filesystem management powered by GlusterFS, delivering 3-way replicated shared volume mobility (/var/contenedores) across Centurion nodes.

🏛 Storage Pool Architecture

graph TD
    Manager["👑 Manager Node (10.10.100.27)"] <-->|Peer Mesh TCP 24007| Worker1["💻 Centurion 1 (10.10.100.25)"]
    Manager <-->|Peer Mesh TCP 24007| Worker2["💻 Centurion 2 (10.10.100.26)"]
    Worker1 <-->|Peer Mesh TCP 24007| Worker2

    SubGraph["🧱 Distributed Replicated Pool (gv_contenedores)"]
    Manager --- SubGraph
    Worker1 --- SubGraph
    Worker2 --- SubGraph

    SubGraph -->|Auto-Mount| Mount["📁 /var/contenedores (All Hosts)"]

🚀 Key Features

  • 3-Way Mirrored Volumes (Replica 3): Automatic distributed volume creation, starting, stopping, and container write-behind cache optimizations (performance.write-behind, flush-behind, stat-prefetch).
  • Trusted Storage Pool Peer Mesh: Live peer discovery, health probing, latency checks, and quorum diagnostics across Centurion worker hosts.
  • Granaries /var/contenedores Auto-Mount: One-click automated mounting across all cluster nodes directly syncing with Gubernator's /etc/fstab management.
  • Ghost Volume Auto-Purge: Intelligent detection and cleanup of stale volume states with 1-click force recreation.
  • Self-Healing & Split-Brain Diagnostics: Real-time heal entry queues, brick health inspection, and manual self-heal triggers.
  • Point-in-Time Snapshots: Instant hot snapshots of GlusterFS distributed volumes with 1-click rollback.

💻 CLI Commands

# Volume Management
gbnt gluster volume ls
gbnt gluster volume create gv_contenedores --replica 3 --brick-dir /data/glusterfs/brick1
gbnt gluster volume start gv_contenedores
gbnt gluster volume stop gv_contenedores
gbnt gluster volume heal gv_contenedores

# Storage Pool Peers
gbnt gluster peer ls
gbnt gluster peer probe 10.10.100.25

# Cluster Auto-Mount
gbnt gluster mount gv_contenedores /var/contenedores --nodes all