Update fastq data archive bucket settings

[reisingerf]

This is to finalise the bucket configuration / setup of our archive buckets before we start the data migration / archival.

Note:

• archive tiers are split in two, for large and smaller files
• versioning is enabled. This may have significant cost implications as all object versions will be charged for a minimal period. It's the only way to guard against accidental deletion though (I think).
• version expiry is based on the minimal charged duration (GLACIER_IA), not on the duration of the bulk of the data (DEEP_ARCHIVE)
• the FileManager role is only given read access, the prod account only has list access and there is no dedicated write access yet

[victorskl]

Related umccr/orcabus#700

[reisingerf]

@alexiswl after our chat the other day, I'm going ahead and apply those changes.