Job deletion is recorded in DB before all files deleted

[abought]

Summary

While updating parameter handling code, I noticed that the database records a job as being archived before all cleanup is complete. Example:
https://github.com/statgen/cloudgene/blob/950b8b8bd5865482e0114f19f70d17fa83a753bf/src/main/java/cloudgene/mapred/api/v2/admin/ArchiveJob.java#L90-L96

In theory, in the event of a temporary communication failure with external workspace, this could lead to files being retained even after the official records say a file was deleted. This would manifest as significantly higher storage usage in the external workspace. Depending on which files were not deleted, it could also have other implications.

Suggested actions

• Change the order of instructions so that the DB archive state is only set after all cleanup operations are complete.
• Audit other delete/archive operations for similar out-of-order behaviors.