Changelog 8/28/2023: Role-based auth and new advanced features

Role-based Authentication

Prequel now supports role-based authentication (or "identity federation") for AWS or GCP destinations, including Redshift, Athena, S3, BigQuery, and Google Cloud Storage! Role-based authentication is a secure authentication method that obviates the need to share credentials to authenticate to a platform. Specifically, this means your customer doesn't need to share credentials (e.g., a username & password) to connect to certain destinations – instead, they may simply grant access directly to a given AWS IAM Role or GCP Service Account. 

As the cloud providers continue to update their recommended integration methods, we're committed to updating our integrations alongside them, and this is our latest improvement towards that goal. 

Note: We will continue to support credential based authentication, though connections that were previously established using credentials or HMAC keys can safely be upgraded without loss of data. Let us know if you'd like specific migration instructions!

Frontend improvements

We're continuing to make usability improvements to the frontend.We've recently added the ability to sync and force refresh specific models on a per-destination basis, as well as the option to edit destination connection details directly in the UI.

Magic link improvements

For teams leveraging the magic link experience, we've recently added some optional fields to the magic link generation flow. The new optional fields are:

• Create as disabled: configure magic link submissions to create destinations as "disabled" so that they can be manually validated before the backfill is started. 
• Redirect after submit: send users to a specified URI after the magic link has been submitted. 

Advanced features

Over the summer, we've been working hard on certain "power-user" features to help our customers as they continue to grow with the Prequel platform. Some of the standouts here are:

• Backfill windows: we've added the ability to configure the max "window" per batch of data. For example, if you were to attempt to backfill a new destination with years of data, you can now force Prequel to work through the backfill in fixed "chunks", to minimize burst resource utilization. Note: this is only not enabled by default. Please reach out to us if you'd like to use this feature. 
• Max concurrent transfers: We've introduced a second mechanism to throttle load on the source for customers with many live destinations. A maximum concurrent transfer limit effectively limits the number of transfers that can run at any given time. The default value is `10`. Note: please reach out to us if you'd like to configure this option for your account. 
• Broader support for source queries: we've introduced the ability to specify a custom query to use when reading data from the source. This has proven useful for certain cases where the query engine requires a specific query format to prune partitions or predicate pushdown and we've expanded our source availability. Note: this is only available on a subset of sources. Please reach out to us if you'd like to use this feature. 

Bug fixes

• Duplicate form submission prevention for destinations and magic links
• Enhanced validation for common fields. Error prone fields like "host", "region", and "bucket_name" now have enhanced field validation.‍
• Improved SSH tunnel connection performance and fixed connection issues.