From 18:46:16 to 18:46:26 UTC, we were alerted to an issue that cased a restart of Tempo Ingestors in the US-East region.
During this time, users may have noticed 404 or 500 errors in their agent logging, potentially resulting in a small amount of discarded Tempo traces for the time the ingesters were not available.
Our Engineers were able to identify the cause and a solution was implemented to resolve the issue. Please contact our support team if you notice any discrepancies of have questions.
A transient error in our infrastructure caused all public probes to report MultiHTTP and Scripted checks as failures for roughly 5 minutes, from 9:55 UTC to 10:00 UTC. The error has been addressed and all probes should now be operating normally.
From 13:49 - 13:54 UTC a deployment to the Frankfurt probe location caused DNS resolution timeouts affecting DNS, HTTP, MultiHTTP, and Scripted checks with failure rates of 20-50% during this time. After a rollback by 13:54 synthetics tests returned to normal.
From 21:30-22:45 UTC, the Seoul public probe location experienced connectivity problems. We observed failure rates of 10-15% across PING, DNS, HTTP, multiHTTP and k6 scripted checks due to connection timeouts. The issue has cleared but we continue to monitor.
The incident was resolved and the affected cell is back to normal. We've identified the root cause of the high latencies and took counter-measures to mitigate the performance problems.