Skip to content

NIP runbooks

Operational playbooks for alerts routed to #nip-ops (platform=nip-bridge).

Wiki base URL: https://wiki.pup.finance/runbooks/nip/
Source of truth: docs/runbooks/nip/ in the monitoring-staging git repo — edit there, run mkdocs build, deploy site/ to the wiki server.

Servers

Droplet Role Tailscale Notes
baasey-nip-server NIP inward 100.94.107.110 /home/inward_dev, port 8009
easypay-nip-integration-prod NIP outward 100.65.109.7 Outward bridge
monitoring-prod Prometheus, Grafana, Loki, archive 100.115.57.112 /opt/monitoring-staging

Runbook index

Scrape & reachability

Alert Severity Runbook
NipNodeExporterDown Critical node-exporter down
NipCadvisorDown Critical cAdvisor down
NipHealthProbeDown Critical /health probe down
NipAppMetricsDown Warning app /metrics down
NipPrometheusTargetDown Warning scrape target down

Host (node_exporter)

Alert Severity Runbook
NipHostCPUHigh Warning host CPU high
NipHostMemoryHigh Warning host memory high
NipHostDiskWarning Warning host disk warning
NipHostDiskCritical Critical host disk critical

Containers (cAdvisor)

Alert Severity Runbook
NipWebContainerDown Critical web container down
NipContainerOOMKilled Critical container OOM
NipContainerRestartLoop Warning container restart loop

Postgres / Redis

Alert Runbook
NipPostgresExporterDown postgres exporter down
NipPostgresConnectionsHigh postgres connections high
NipPostgresSlowQueries postgres slow queries
NipPostgresDeadlocks postgres deadlocks
NipRedisExporterDown redis exporter down
NipRedisMemoryHigh redis memory high
NipRedisEvictions redis evictions
NipRedisHitRateLow redis hit rate low

Business metrics (requires app /metrics)

Alert Runbook
NipAllWorkersBusy all workers busy
NipTransferFailureSpike transfer failure spike
NipDependencySlow dependency slow

Spaces archival (monitoring-prod)

Alert Runbook
NipSpacesArchiveFailed spaces archive failed
NipSpacesArchiveStale spaces archive stale
NipSpacesArchiveMetricsMissing spaces archive failed

After editing runbooks

  1. Commit changes in monitoring-staging.
  2. On wiki host: mkdocs build → publish site/ to wiki.pup.finance.
  3. On monitoring-prod: curl -X POST http://127.0.0.1:9091/-/reload (if alert rules changed).