Skip to content

Operations

Running qkt in production: deployment, monitoring, troubleshooting.

  • Deploy with Docker

    The full stack walkthrough: qkt + mt5-gateway via docker-compose.yml.

    Deploy with Docker

  • Production deploy (qkt-prod)

    Runbook for the managed Dokploy/Swarm host: release cadence, rollback, "what's deployed".

    Production deploy

  • Monitoring

    Liveness, equity, P&L, alerts. What to scrape and what to page on.

    Monitoring

  • Metrics

    The JSON endpoints qkt exposes, plus how to write a sidecar that reformats them for Prometheus, Datadog, or any other monitoring stack.

    Metrics

  • State backup

    Backing up and restoring the qkt-state directory — cadence and mechanism.

    State backup

  • Logging

    MDC keys, console + file patterns, logback overrides, JSON output.

    Logging

  • Troubleshooting

    Symptom → cause → fix matrix. Use when something's wrong.

    Troubleshooting

Not yet covered

  • Capacity planning — strategies-per-daemon ceiling, port allocation, JVM heap sizing
  • Alerting playbook — concrete runbooks per alert type

Pages stubbed for a future operations-depth phase. If you need one urgently, open an issue.