Summary/Problem Statement

The rhsm-subscriptions-egress job has been failing in production for an extended period (SWATCH-4024) but is no longer used upstream. Its original use case has been replaced by Floorist queries for Telesense, and our Snowflake integration will take a different implementation path that doesn't require reading from these S3 buckets. When this job fails, it reports failures to the #alerts-swatch Slack channel, which creates noise and buries real critical alerts.

Acceptance Criteria

• Add delete: true to rhsm-subscriptions-egress resource template targets in deploy.yml for both stage and prod namespaces
• Add delete: true to rhsm-subscriptions-egress-post-deploy-tests target in post-deploy-tests.yaml
• Add delete: true to any rhsm-subscriptions-egress references in deploy-perf.yml if they exist
• Verify that standalone egress deployments are removed from stage and prod environments
• Confirm that swatch-tally service continues to function normally (including database cleanup jobs)

Done Criteria

• All standalone egress job deployments are disabled across all environments
• No standalone egress job containers are running in stage or production
• Main swatch-tally service continues to function normally (including database cleanup jobs)
• No S3 upload failures are occurring from disabled egress job
• Evaluate is SWATCH-4024 is resolved by these changes and if so, it can be closed out

Implementation Notes

• Use delete: true on resource template targets in app-interface files to safely disable deployments
• Main deployment: data/services/insights/rhsm/deploy.yml
• Post-deploy tests: data/services/insights/rhsm/post-deploy-tests.yaml
• The swatch-tally ClowdApp references EGRESS_IMAGE parameters, but this is only for reusing the egress container image as a PostgreSQL utility container for database cleanup jobs - do not modify swatch-tally
• Deploy changes to stage environment first, then after confirming the next day's deployment cycle, open a separate MR to apply the same changes to production
• The actual egress functionality that exports data to S3 is deployed as separate standalone deployments, not as part of swatch-tally
• The egress post-deploy tests (rhsm-subscriptions-egress-post-deploy-tests) run automatically after successful egress deployments and publish to rhsm-subscriptions-egress-stage-post-deploy-tests-success-channel - these will be disabled along with the main deployment
• This is a safe disable operation that preserves all configurations for potential rollback
• This is Phase 1 of the egress decommissioning - subsequent cards will be created for complete cleanup (repository cleanup, documentation updates, etc.)