Status

1 component reporting degraded performance.

Component-level health with ninety days of history. Capture and the console are tracked separately, because one being slow does not mean the other is.

DegradedUpdated continuously · demonstration data
  • 99.98%Capture availability · 90d
  • 99.97%Console availability · 90d
  • 1.8msMedian collector overhead
  • 1Open incidents
Components · last 90 daysOperationalDegradedOutage
  • Edge collectors — USiad1, sfo1, ord199.99%operational
  • Edge collectors — EUfra1, lhr1, cdg199.98%operational
  • Edge collectors — APsyd1, sin1100%operational
  • Console and APIapp.flintwake.com, api.flintwake.com99.97%operational
  • Webhook deliveryOutbound event fan-out99.91%degraded
  • Hosted documentationDocs, reference and playground100%operational
  • Data exportS3, BigQuery, Snowflake destinations99.95%operational

Incident history

  • MinorMonitoring16 September 2026 · Ongoing · 41 minutes

    Elevated webhook delivery latency

    • 12:04 UTC

      Delivery latency is elevated for endpoints in the EU fan-out pool. Median delivery is 4.1s against a normal 180ms. No events have been lost; retries are proceeding normally.

    • 11:48 UTC

      We have identified a saturated queue partition following an unusually large replay job and are rebalancing it.

    • 11:23 UTC

      We are investigating reports of slow webhook delivery in the EU region.

  • MajorResolved17 July 2026 · 1 hour 12 minutes

    Webhook delivery unavailable in eu-west

  • MinorResolved28 June 2026 · 34 minutes

    Console read latency in us-east

How we report

The status page and the changelog agree.

If a week was bad, it appears in both places and they link to each other. Incident write-ups are published within five working days, whether or not anyone asked.

  • Subscribe

    Register a webhook endpoint for status.changed, or write to support@flintwake.com to be added to the incident mailing list.

  • Severity

    Major means capture stopped or data was at risk. Minor means something was slow or a secondary surface was unavailable.

  • History

    Ninety days on this page; the full record is kept indefinitely and available on request.

Reliability is a claim until it has a history.

Ninety days of component-level uptime, the incidents that broke it and what changed afterwards.