The describe_time_to_live api call is heavily throttled, we don't want to have to call it for every dynamodb table. This pr separates out the describe ttl call into a separate function, and only calls it on tables who have at least an average of 100tps over the course of a month. We can tune this heuristic later on
The describe_time_to_live api call is heavily throttled, we don't want to have to call it for every dynamodb table. This pr separates out the describe ttl call into a separate function, and only calls it on tables who have at least an average of 100tps over the course of a month. We can tune this heuristic later on
Before
After