Feature/job reliability

[rosemcc]

This pull request introduces enhancements and refactoring to support ActiveScale integration and improve archive job tracking. Key changes:

• ActiveScale integration
• Revised archive submission/job tracking model
• Manifest model and related code have been removed

ActiveScale Integration

• Added S3/ActiveScale integration module for handling initialisation of Activescale S3 client, as well as generic S3 operations.
• Added a shell script scripts/check_activescale.sh for verifying ActiveScale connectivity and permissions, including environment loading and proxy handling.
• Added boto3 and types-boto3-s3 as dependencies for ActiveScale S3 API access in pyproject.toml.

Configuration and Environment

• New environment variables for ActiveScale configuration (ACTIVESCALE_HOSTNAME, ACTIVESCALE_REGION, etc.) in .env.development and .env.production.
• Extended the Settings class in src/config.py to include ActiveScale configuration and secrets, using SecretStr for sensitive values.

Archive Submission Model Refactor

• Refactored the ArchiveSubmission model to track job lifecycle stages using a new JobStage enum, with explicit timestamps and retry/cleanup tracking. Removed legacy fields and manifest relationship.
• Updated the SubmissionResponse model to reflect the new job stage and timestamp fields, and removed obsolete fields.
• Updated the frontend api client and functionality to work with the new submission model.

Manifest Model Removal

• Removed the Manifest model and related code, since for now there's no reason to store the list of files in the db