Details
-
Improvement
-
Resolution: Done
-
Major
-
Cheshire-Cat
Description
Although we attempt to ensure that we're not duplicating code as much as possible (we share a common file based lockfile implementation for archive/staging), we currently must have native implementations for cloud lockfiles due to differences in behavior.
With that being said, there's a fair chance we can share a subset of the core logic and use a common implementation between object store clients. This however depends on the consistency of Azure blob stores, which was of prime consideration when we implementation the AWS lockfile.