After we refactored the blocktime function, it resulted in skipping cache for every invalid key as we stopped caching invalid keys in redis to prevent stale data, whilst this was good it resulted in excessive ceramic node calls if a blocktime wasn't returned (null).
Explanation of the solution
Added a second layer caching solution to the getBlockTime function, and only calling a retrieve on each unique commit/txHash
Description of the Problem / Feature
Explanation of the solution