Network status
Updated 22 September 2026, 21:05 UTC. Refreshes every minute. Times on this page are UTC.
Scheduled maintenance
Planned failover of the primary database to its replica. The API and panel stay available; requests may see up to a few seconds of added latency during the switch. Message routing, webhooks and deposits are not affected.
HistoryPast incidents
Each entry keeps the full timeline as it was posted. Durations are from first report to resolution. Maintenance windows are listed with the incidents.
We are seeing a lower delivery rate than usual to one operator in Middle East and Africa. Other operators in the region are unaffected.
The route to this operator is filtering a share of traffic. We are swapping it for an alternative route.
Traffic to the affected operator is now on the alternative route and delivery rates have recovered. We are monitoring.
Delivery rates to the operator have been normal for the past hour. Incident resolved.
Webhook events are being delivered with a delay. No events are lost; they are queued and will be delivered in order.
The webhook dispatcher is working through a backlog caused by a slow downstream endpoint saturating a worker pool. Isolation per endpoint has been enforced.
The backlog is clearing and delivery delay is back under one minute for most endpoints.
All queued events have been delivered. Incident resolved.
Delivery receipts from one operator in North America are arriving late. Messages are delivering normally; statuses may stay at sent longer than usual.
The operator confirmed a backlog on its receipt gateway. No action is needed on your side; statuses will update as the backlog clears.
Receipts are flowing again and the backlog is being processed. Webhooks are being emitted for each late receipt.
All receipts have been processed. Statuses and webhooks are up to date.
Planned failover of the primary database to its replica. The API and panel stay available; requests may see up to a few seconds of added latency during the switch. Message routing, webhooks and deposits are not affected.
Completed as planned. No customer impact was recorded.
Messages to Middle East and Africa are taking longer than usual to be accepted by carriers. Delivery is not impacted.
A congested interconnection has been identified. We are rebalancing traffic across the available links.
Latency is back to normal levels for the region. We continue to monitor.
Latency has remained normal. Incident resolved.
Planned failover of the primary database to its replica. The API and panel stay available; requests may see up to a few seconds of added latency during the switch. Message routing, webhooks and deposits are not affected.
Completed as planned. No customer impact was recorded.
Messages to Middle East and Africa are taking longer than usual to be accepted by carriers. Delivery is not impacted.
A congested interconnection has been identified. We are rebalancing traffic across the available links.
Latency is back to normal levels for the region. We continue to monitor.
Latency has remained normal. Incident resolved.
A share of requests to the API are returning 5xx errors. Requests that failed were not sent and were not billed; retry with the same Idempotency-Key.
A faulty node in the API pool has been identified and removed from rotation. Error rates are dropping.
Error rates have returned to baseline. We are monitoring while the node is rebuilt.
The API has been stable for the past hour. Incident resolved.
Delivery receipts from one operator in North America are arriving late. Messages are delivering normally; statuses may stay at sent longer than usual.
The operator confirmed a backlog on its receipt gateway. No action is needed on your side; statuses will update as the backlog clears.
Receipts are flowing again and the backlog is being processed. Webhooks are being emitted for each late receipt.
All receipts have been processed. Statuses and webhooks are up to date.
Messages to Middle East and Africa are taking longer than usual to be accepted by carriers. Delivery is not impacted.
A congested interconnection has been identified. We are rebalancing traffic across the available links.
Latency is back to normal levels for the region. We continue to monitor.
Latency has remained normal. Incident resolved.
We are seeing a lower delivery rate than usual to one operator in North America. Other operators in the region are unaffected.
The route to this operator is filtering a share of traffic. We are swapping it for an alternative route.
Traffic to the affected operator is now on the alternative route and delivery rates have recovered. We are monitoring.
Delivery rates to the operator have been normal for the past hour. Incident resolved.
Webhook events are being delivered with a delay. No events are lost; they are queued and will be delivered in order.
The webhook dispatcher is working through a backlog caused by a slow downstream endpoint saturating a worker pool. Isolation per endpoint has been enforced.
The backlog is clearing and delivery delay is back under one minute for most endpoints.
All queued events have been delivered. Incident resolved.
Planned failover of the primary database to its replica. The API and panel stay available; requests may see up to a few seconds of added latency during the switch. Message routing, webhooks and deposits are not affected.
Completed as planned. No customer impact was recorded.
Some keys are receiving 429 responses below their limit. Retrying after the Retry-After header succeeds.
A rate-limiter node had a stale configuration. It has been reloaded.
No incorrect 429 responses observed since the reload. We are monitoring.
Incident resolved.
API responses are slower than usual. Requests complete successfully.
The cause is a slow query on the message index after a schema change. A fix is being rolled out.
Latency has returned to normal after the fix. We are monitoring.
Latency has remained normal. Incident resolved.
Some users see an error when signing in to the panel. Sending through the API is not affected.
A session store node has been restarted after a memory issue.
Sign-in is working normally. We are monitoring.
Incident resolved.
Generating a new deposit address is slower than usual. Existing addresses work normally.
The address service was waiting on a rate-limited upstream. The dependency has been removed.
Address generation is back to normal speed. We are monitoring.
Incident resolved.
We are seeing a lower delivery rate than usual to one operator in Europe. Other operators in the region are unaffected.
The route to this operator is filtering a share of traffic. We are swapping it for an alternative route.
Traffic to the affected operator is now on the alternative route and delivery rates have recovered. We are monitoring.
Delivery rates to the operator have been normal for the past hour. Incident resolved.
Planned failover of the primary database to its replica. The API and panel stay available; requests may see up to a few seconds of added latency during the switch. Message routing, webhooks and deposits are not affected.
Completed as planned. No customer impact was recorded.
Planned failover of the primary database to its replica. The API and panel stay available; requests may see up to a few seconds of added latency during the switch. Message routing, webhooks and deposits are not affected.
Completed as planned. No customer impact was recorded.
We are seeing a lower delivery rate than usual to one operator in Asia-Pacific. Other operators in the region are unaffected.
The route to this operator is filtering a share of traffic. We are swapping it for an alternative route.
Traffic to the affected operator is now on the alternative route and delivery rates have recovered. We are monitoring.
Delivery rates to the operator have been normal for the past hour. Incident resolved.
Planned failover of the primary database to its replica. The API and panel stay available; requests may see up to a few seconds of added latency during the switch. Message routing, webhooks and deposits are not affected.
Completed as planned. No customer impact was recorded.
Generating a new deposit address is slower than usual. Existing addresses work normally.
The address service was waiting on a rate-limited upstream. The dependency has been removed.
Address generation is back to normal speed. We are monitoring.
Incident resolved.
Webhook events are being delivered with a delay. No events are lost; they are queued and will be delivered in order.
The webhook dispatcher is working through a backlog caused by a slow downstream endpoint saturating a worker pool. Isolation per endpoint has been enforced.
The backlog is clearing and delivery delay is back under one minute for most endpoints.
All queued events have been delivered. Incident resolved.
History starts 16 March 2026, when this page went live. Degraded performance counts half in uptime; outages count in full.
