Skip to content

Latest commit

 

History

History
49 lines (38 loc) · 1.47 KB

File metadata and controls

49 lines (38 loc) · 1.47 KB
title Operations
description Day-2 operator handbook.

Operations

Day-2 operator content: how to keep pg_hardstorage healthy, monitor it, plan capacity, and recover from incidents.

Pages

Handbook

Observability

  • Monitoring — Prometheus scrape, OpenTelemetry tracing, structured-JSON logs, and the pg_hardstorage_* metric namespace.
  • Alerting recipes — drop-in PromQL / Alertmanager rules for backup overdue, WAL silence, KEK unreachable, scrub findings, and more.

Incident response

  • Incident response — symptom → action mapping with cross-links to the R1–R7 runbooks.

Planning

  • Capacity planningpg_hardstorage capacity report with 30/90/365-day projection.
  • Cost reporting — per-deployment and per-tenant repo-cost exports.
  • SLO as code — declarative RPO/RTO targets and the alert wiring.
  • Scaling to large fleets — staying fast and bounded across thousands of deployments: the job-concurrency cap, agent poll jitter, sharded audit chains, and the deployment index.