Previous incidents

May 2024
May 14, 2024
1 incident

Evaluations and Configs Were Unavailable

Resolved May 14 at 05:16pm EDT

The ability to perform javascript-client evaluations and fetch configurations were down.
Many configuration loads would have been satisfied via our CDN during this time so only some customers were affected.

We've investigated and fixed a few issues to prevent this in the future
* improved anti-affinity rules to spread API service instances into more availability zones
* installed a missing set of Load balancer health checks for the API

April 2024
Apr 16, 2024
1 incident

New evaluations failing

Resolved Apr 16 at 09:24pm EDT

50 seconds of downtime for the evaluation endpoint. New visitors to JS clients received fallback values. Existing visitors got cached values.

Apr 13, 2024
1 incident

New evaluations failing

Resolved Apr 13 at 09:07pm EDT

30 seconds of downtime for the evaluation endpoint. New visitors to JS clients received fallback values. Existing visitors got cached values.

March 2024
Mar 27, 2024
1 incident

Intermittent API Issues

Resolved Mar 27 at 03:45pm EDT

3:18 - We're investigating API stability issues.

3:43 - We've resolved the issues.

We had issues scheduling new pods due to an overly restrictive anti-affinity policy. This exacerbated a deployment issue.