I want to test the claim that Microsoft has for Azure Table Storage Partitions. They say that they get 500 IOPS per partition... but I want to see for myself and also look at how the storage client library will react to an overloaded partition.
Since I've seen various issues with the servicePointManager and other items regarding this, I'd like to know if anyone else has been able to reach 500 IOPS for a partition and how they did it. Can there be more than 500 IOPS? What is the real world range?