Expected Lambda Throttling Behavior and Performance Benchmarks #181
Description
I noticed that there's consistent throttling of the IcebergMetadataWriterFunction (>100/5m) and LakeWriterAlertsFunction (>350/5m) lambdas and was wondering if this is expected behavior or if there's some tuning that should be done to increase the processing throughput of the lambda (e.g. increase RAM). The documentation states that there should only be one concurrent execution of the lambdas, I'm wondering in addition to if the throttling from this will cause issues, are there concerns about Matano scaling to be able to keep up with the increase in capacity? Are there any benchmarks or known throughput limits that you could share so we can ensure we don't overload Matano?