Make use of batch and etags in Azure Table API to ensure data integrity.

[MattPitlyk]

With the current AzureTableStorage implementation, there is the potential for data corruption either by 1) one thread doing multiple operations and failing on a later step, which leaves the data in an invalid state or 2) two threads work on the same data at interweaving intervals.

The batch mechanism in the Azure Table API will fix 1.
Making use of etags will help with 2.