500's on OfferRequest

We’ve seen a significant improvement in search latency after updating our scaling config, returning to pre-incident levels.

We’re continuing to monitor 500s being returned on offer requests and will provide an updates as we work through the individual cases tomorrow.

We’ve updated our scaling policy and are currently monitoring search latency.

We’re seeing an increased number of offer_requests, resulting in an increase in latency for P99 of offer requests. For a subset of these requests this is resulting in a 500 error code being returned.

We’re reverting a scaling config change made earlier in the day to reduce this latency and error rates.

We are currently investigating some of 500's being returned on the OfferRequest endpoint.

The team is currently investigating impact and will provide more details shortly.