Cloudflare D1 was down / degraded on July 3, 2026 (resolved)

Update July 6, 2026: Cloudflare has marked this incident resolved. The official incident page says it was resolved on July 3, 2026 at 20:30 UTC, and Cloudflare's current status summary lists both D1 and Durable Objects as operational.

If your Cloudflare D1-backed app was throwing errors on July 3, 2026, you were not alone. At about 16:18 UTC, Cloudflare's status page listed an active incident called Durable Objects increased errors in ENAM, and the incident listed D1 as one of the affected services.

This was worth calling out because the problem did not look like a standalone D1-only outage. Cloudflare said it was investigating increased errors for Durable Objects and downstream services in Eastern North America, and Cloudflare's own D1 limits documentation says each individual D1 database is backed by a Durable Object. That made the Durable Objects incident a very plausible reason for D1 apps to see failures, slow queries, or intermittent database errors.

What Cloudflare is reporting

During the incident, the official Cloudflare Status page showed a minor service outage, and the status API showed D1 and Durable Objects under degraded performance for this incident. The affected service list also includes Realtime SFU, RealtimeKit, and Workflows.

Cloudflare later marked the incident resolved, with D1 and Durable Objects returning to operational status. Cloudflare did not publish a detailed root cause in the incident updates.

What to do if your app uses D1

If you were debugging errors from July 3, check the Cloudflare status page and your Worker logs. If failures started around 16:13 UTC and your app depends on D1, Durable Objects, or Workflows, this incident was a likely suspect.

For production apps, make sure failed database calls are logged clearly and that any retries use backoff. If requests are not idempotent, be careful about blindly retrying writes, because the first attempt may have partially completed even when the user saw an error.

The incident is resolved, but keeping clear D1 and Worker logs is still useful for separating platform issues from local application bugs the next time an outage happens.