Admin Guide

This guide is for operators and platform engineers running ShrikeDB. It covers startup, configuration, user administration, security, HA, observability, and operational validation.

Operator Mental Model

process
  server binary
    |
    +-- gRPC data plane
    +-- HTTP convenience API
    +-- admin health/traces/metrics
    +-- optional HA peer traffic

state
  storage.data_dir
    |
    +-- segments and WAL
    +-- graph projection journals
    +-- HA ledgers and repair queues
    +-- identity records in system KV namespaces

Deployment preparation

  • Run release binaries.

  • Use explicit config files rather than relying only on defaults.

  • Set distinct bind and advertise addresses for HA nodes.

  • Set SHRIKEDB_CLUSTER_NODE_TOKEN when HA is enabled.

  • Use multi-tenant auth when multiple principals need different permissions.

  • Create tenant-scoped roles for analysts, ingestion jobs, lakehouse reviewers, and admins.

  • Capture health, trace, and benchmark artifacts during validation.

  • Retain source ontology files and input-data versions with the deployment records.

Admin Tasks

Task Page

Install or start a node

Prepare a managed server

Configure server behavior

Configuration

Manage tenants, users, roles, policies, and API keys

User Administration

Understand authentication and authorization enforcement

Security Model

Evaluate clustered operation

HA deployment status and configuration

Observe health, metrics, and traces

Observability