Large table drop causes issues when interrupted

Issue:

1. Create a table with large number of partitions (e.g 10K partitions)

2. Drop this table (This will take lot longer time in s3 & other calls)

3. Interrupt #2. In beeline try "ctl+c" or cancel query in hue.

4. Since interrupt handling has issues in this codepath, it doesn't kill the query.

Other statements will start waiting from this point onwards, as the lockManager doesn't release the locks completely.