Skip to content

Monitoring

Akasha exposes Prometheus metrics at /metrics for integration with Grafana, Datadog, or any Prometheus-compatible scraper.

Prometheus Configuration

# prometheus.yml
scrape_configs:
  - job_name: 'akasha'
    scheme: https
    tls_config:
      insecure_skip_verify: true
    static_configs:
      - targets:
          - 'akasha-01:7777'
          - 'akasha-02:7777'
          - 'akasha-03:7777'
    scrape_interval: 15s

Available Metrics

Metric Type Description
akasha_records_total Gauge Total records in the store
akasha_agents_total Gauge Registered agents
akasha_pheromones_active Gauge Active pheromone trails
akasha_memory_records{layer} Gauge Records per memory layer
akasha_cluster_nodes{status} Gauge Cluster nodes by status
akasha_crdt_tracked_paths Gauge CRDT tracked paths
akasha_crdt_pending_deltas Gauge Pending sync deltas
akasha_raft_commit_index Gauge Raft commit index
akasha_raft_term Gauge Current Raft term
akasha_raft_is_leader Gauge Whether this node is the leader
akasha_nidra_is_leader Gauge Whether this node runs Nidra

Grafana Dashboard

Import the ready-made Grafana dashboard:

  1. Download deploy/grafana-dashboard.json
  2. In Grafana → Dashboards → Import → Upload JSON
  3. Select your Prometheus data source
  4. Done! 12 panels showing all Akasha metrics.

The dashboard includes:

  • Stat panels: cluster health, records, agents, pheromones, CRDT sync, leader status
  • Time series: records over time, memory layers (stacked), cluster nodes, CRDT deltas, Raft consensus, pheromone activity

Manual Check

curl -sk https://localhost:7777/metrics | head -20

# HELP akasha_records_total Total records in the store.
# TYPE akasha_records_total gauge
akasha_records_total 569
# HELP akasha_memory_records Records per memory layer.
# TYPE akasha_memory_records gauge
akasha_memory_records{layer="working"} 38
akasha_memory_records{layer="episodic"} 467
akasha_memory_records{layer="semantic"} 50
akasha_memory_records{layer="procedural"} 2