All systems operational.
Live service availability and recent maintenance history.
Updated 9 April 2026 at 14:02 UTC
Systems
OperationalDegraded
API
Public REST and streaming endpoints, all regions.
90 days ago99.98% uptimeToday
Dashboard
Web console, project settings and billing views.
90 days ago99.99% uptimeToday
Edge network
Cache and routing tier in front of every origin.
90 days ago99.91% uptimeToday
Webhook delivery
Outbound event delivery and retry scheduling.
90 days ago99.62% uptimeToday
Authentication
Sign-in, token issuance and key rotation.
90 days ago100.00% uptimeToday
Object storage
Bucket reads, writes and lifecycle rules.
90 days ago99.97% uptimeToday
Past incidents
Times shown in UTC7 April 2026
Delayed webhook delivery in eu-west
Monitoring
- 16:41
- The backlog has drained and new events are being sent within the usual two seconds. We are leaving this open overnight while the retry queue works through the rest.
- 15:20
- Added four workers to the eu-west delivery pool. The oldest queued event is now 9 minutes behind, down from 41 minutes.
- 14:58
- We are investigating delayed outbound webhooks for projects in eu-west. Events are being accepted and stored; delivery is running behind. No events have been dropped.
29 March 2026
Scheduled database upgrade — completed
Resolved
- 04:35
- The upgrade finished ahead of the window and all replicas are serving reads.
- 03:12
- Primary failover completed in 38 seconds. Writes were rejected during that period and clients retrying with backoff succeeded.
- 02:00
- Maintenance has begun as announced. Expect brief write errors while the primary is promoted.
18 March 2026
Elevated errors on cached edge responses
Resolved
- 11:26
- Error rates have been at baseline for 90 minutes. A validation step now blocks that configuration from reaching the rollout pipeline.
- 09:44
- A cache configuration pushed at 09:31 was rolled back. Roughly 0.6% of requests in three locations returned 502 for 13 minutes.
2 March 2026
Sign-in failures for accounts using hardware keys
Resolved
- 21:08
- All affected accounts can sign in. Sessions created before 19:50 were unaffected and no credentials needed resetting.
- 20:15
- Identified an expired signing certificate in the assertion verifier. A replacement is being deployed region by region.
- 19:52
- We are investigating reports that hardware key sign-in is rejected. Password and recovery-code sign-in is unaffected.
99.94%Uptime, last 90 days
99.99%Uptime, last 30 days
142 msAPI response, 24 h median
One email when an incident is opened and one when it closes. Maintenance windows are announced two days in advance.
Scheduled maintenance
Storage metadata migration
Sunday 26 April 2026, 01:00 – 04:00 UTC
Bucket listings may take up to a minute to reflect new objects during the window; reads and writes continue as normal.