Berserk Docs

Ingest & Operations

Ingestion, cluster administration, and configuration for Berserk

  • Self-Hosting Berserk — start here: an overview of the configuration surface (secrets, user integration, network ingress) with links into the detail below
  • Hardware Sizing — how many cores, how much memory, disk and object storage to plan for, from your ingest volume and retention, with a calculator
  • Dependencies — external dependencies and environment-specific deployment (default, EKS)
  • Storage — local disk configuration for caching and temporary storage
  • Ingestion — how data flows into Berserk, the ingest contract, and per-shipper configuration (OpenTelemetry Collector, Fluent Bit)
  • Cluster Admin — services and Helm configuration
  • Observability — monitoring Berserk itself: the Prometheus /metrics endpoint on every pod, and exporting Berserk's own telemetry over OTLP
  • Running Queries — bzrk ps and bzrk kill: what the cluster is executing, whether a query is slow or stuck, and where each one came from
  • Query Usage Reporting — the anonymized per-query usage record: what it contains, why it cannot contain your data, and where it goes
  • Service Metrics — auto-generated reference for all service metrics
  • Security — locking the cluster down to its known peers with Network Policies, and serving the UI under a strict Content Security Policy with a CSP nonce
  • API Introduction — how the gRPC streaming and HTTP REST protocols work, plus endpoint reference
  • Client Libraries — official SDKs for Python, Rust, Go, and Node.js
  • Third-Party Licenses — attribution for bundled artifacts outside the package-manager dependency graph

Self-Hosting Berserk

The configuration surface of a self-hosted Berserk cluster — secrets, user integration, and network ingress — with links to the detailed guides.