Monitoring, backup & recovery

Observe the runtime, retain recoverable backups and rehearse restoration.

Monitoring signals

Host: CPU, memory, filesystem capacity, disk latency and network errors.
Database: connection count, replication lag, slow queries and storage growth.
Object storage: request errors, capacity, latency and lifecycle failures.
Application: request rate, latency percentiles, error rate and queue depth.
CI: queue time, job duration and failure rate.

Backup workflow

Define recovery point objective (RPO) and recovery time objective (RTO).
Back up database and repository data consistently.
Copy backups to a separate failure domain.
Encrypt backups and restrict restore permissions.
Apply retention and lifecycle rules.
Run restore drills and record measured recovery time.

PostgreSQL dump example

pg_dump --format=custom --no-owner \\
  --file=woww-db.dump "$DATABASE_URL"

pg_restore --list woww-db.dump

A backup job is not proven until a restore has completed and the restored application has been verified.

Scope note: WOWW 0.2.0 is a framework package. Hosted Git collaboration, account management, Copilot plans, hosted runners, public REST/GraphQL services and a woww/core server image are not included. The related guidance below is architecture or integration advice, not a claim that those services are live.

← Documentation home