Repository navigation
ECS Deploy: Fix resource prefix issue #1146
Description
Is your feature request related to a problem?
There is a mismatch in the AWS_RESOURCE_PREFIX and cluster-name during the staging ECS deploy workflow, causing failures and ambiguity due to multiple clusters.
Describe the solution you'd like
- Correct the
AWS_RESOURCE_PREFIXand cluster-name secrets to target the appropriate staging ECS cluster. - Verify that the deploy to staging ECS is successful.
Original issue
Context
During KMS v2 staging ECS validation, the deploy workflow failed on an AWS_RESOURCE_PREFIX / cluster-name mismatch — the workflow was targeting platform-staging names rather than the correct staging ECS cluster names.
There are two clusters (AI-platform-staging and kaapi-staging) with different DBs and secrets, which contributed to the ambiguity.
Acceptance Criteria
- Correct the
AWS_RESOURCE_PREFIX/ cluster-name secrets so the deploy workflow targets the correct staging ECS cluster. - Verify a clean deploy to staging ECS.
Metadata
Metadata
Assignees
Labels
Type
Projects
- Status
Activity
[-]Fix AWS_RESOURCE_PREFIX / cluster-name mismatch in staging ECS deploy workflow secrets[/-][+]ECS Deploy: Fix resource prefix issue[/+]Ayush8923 commentedon Oct 17, 2026
Just a note: we no longer need to make name changes based on the existing ECS staging cluster. Since the non-prod account is now completely separate from the production account, we should update the configuration to use the new non-prod ECS cluster name instead of the old one. We can also consider removing the old ECS cluster if it’s no longer needed.