Postmortem -
Read details
Jul 9, 15:48 EDT
Resolved -
From the time of 3:44-4:18pm EST a subset of customers experienced elevated latency of flag checks, event processing, and the API. The root cause was identified, the issue has been resolved.
At around 4:30 PM we received a large number of events that overwhelmed our caching system. This is a hot path in our system that increased API latency substantially for a subset of customers. We're putting some additional mitigations in place now to further prevent similar issues.
Jul 8, 16:00 EDT