Improve storage manager to handle concurrent rollback requests from different buckets concurrently

Currently, storage manager can process rollback request for one bucket at a time. When there are rollback requests from multiple buckets, the rollback processing is serialised at indexer. This can block indexer main loop and affect cluster wide operations like rebalance (as seen in MB-37105)

The purpose of this improvement is to make storage manager handle rollback requests from multiple buckets concurrently so that indexer main loop is not blocked