VAT Search Tool

Monitoring Hub

One-click access to every observability tool in the VAT pipeline stack.

Dashboards

16-panel VAT pipeline overview — HMRC latency, scan progress, hit rate, Meilisearch, error logs.

Task Queue

Live Celery worker status, tasks per second, active / failed / retried jobs.

Alerts & Metrics

Raw metrics explorer, 9 alert rules status, PromQL query interface.

Scrape Endpoint

Live Prometheus text-format output from this Django app — what Prometheus scrapes every 15s.

View in Prometheus →

HMRCApiHighErrorRate >5% errors for 5 min critical

ScanProgressStalled Scan position unchanged for 1 h critical

HMRCApiRateLimitSpike >0.5 rate-limits/s for 2 min warning

HMRCApiHighLatency p95 latency >5 s for 5 min warning

ScanQueueEmpty No jobs dispatched in 15 min warning

ValidationExceptionRateHigh >0.1 exceptions/s for 3 min warning

MeilisearchIndexingFailures >0.05 failures/s for 5 min warning

TaskDurationP95High p95 task duration >15 s warning

ValidHitRateTooLow Valid rate <2% for 30 min info

Open Grafana Explore →

All errors (worker + web)

{service=~"web|worker|beat"} | json | level="error"

HMRC rate-limit (429) events

{service="worker"} | json | event="hmrc_rate_limit"

OAuth2 token refresh failures

{service="worker"} | json | event="token_refresh_fatal"

Meilisearch indexing failures

{service="worker"} | json | event="meilisearch_index_failure"

VAT status transitions (invalid → valid)

{service="worker"} | json | event="vat_status_change"

Scan discovery batches

{service="worker"} | json | event="discovery_batch_complete"