When the indexing-service ECS tasks are at max CPU load, they can produce redis-related errors due to timeouts. However, it is surprising that one of them is much more frequent than others:
2025-07-29T20:48:45.425Z ERROR providerindex providerindex/providerindex.go:242 setting no results expirable: error accessing redis: context deadline exceeded
We should investigate this issue to determine why this error is more likely than others under high CPU loads, as it might be a fixable bug.