Data Processing Delays

We have identified the bottleneck in our ingestion pipeline and applied a preliminary fix. Event ingestion via the HTTP endpoint is recovering. The real-time HTTP endpoint has fully caught up, while the batch endpoint is currently in the process of recovering.

To prioritize real-time traffic, warehouse import has been paused. As a result, imports are currently delayed by up to 7 hours. We’re deploying another fix to improve performance and support full recovery.

EU customers remain unaffected, and no data loss is expected.

We’re still looking into the issue to better understand and resolve it.

We are continuing to work on a fix for this issue.

We’ve deployed the hotfix, and systems are now recovering.

There is a backlog of several hours of data to process. We’re closely monitoring progress and adjusting configurations as needed.

Thank you for your patience.

We believe we've identified the root cause. We're implementing and deploying a hotfix to confirm this and restore normal data processing.

We're currently experiencing delays in our real-time ingestion and evaluation systems. This issue began around 5:00 PM PST.

Our ingestion systems are operational but processing more slowly than normal. Customers using the HTTP endpoints in the US region are affected. Impacted users may not see events from the last 40–60 minutes reflected in charts. Event streaming out is also delayed.

Our evaluation systems are operational but experiencing higher than normal latency.

No data loss is expected. There is no impact to customers in the EU region.

We're actively investigating and will share an update within the next hour, or sooner if possible.