Although we've implemented a workaround to disable retry logic in AWS, the retry logic is still active in Azure. There is a possibility that certain statuses could still be in a "retryable" state in the Status topic, including with retry_after values. As a result, we plan to remove the column and its associated Enum in the next deployment after SWATCH-3109 in the card SWATCH-3129.

We no longer want to retry usages that have failed, instead missing usage will be caught up using the current_total of the tally_snapshots on the following usages that are submitted(no change, existing functionality)

Todo:

• Deprecate 'retryable' status
• Drop 'retry_after' column on billable_usage_remittance - Will be done in SWATCH-3129 to avoid any null pointers.
• Stop sending 'retryable' statuses on billable-usage status topic from swatch-produzer-aws/azure. These can just be failed instead
• Remove cron job and processRetries endpoint
• Remove BillableUsageDeadLetterConsumer since that is stale anyways

Done:

• When a new billable-usage is created and there is no subscription available and it is within the 96hour azure window, the usage should get marked as 'failed' instead of 'retryable'
• When a billable-usage can't be sent to AWS due to a rate_limit exception. The usage should be 'failed' instead of 'retryable'
• Cron job removed for processRetries