Cognito configuration ignored on redeploy when using existing User Pool

[amitkumawat-node]

Summary

On initial deployment, a new Cognito User Pool is created as expected when no environment variables are provided.

However, on subsequent deployments, even when explicitly providing:

• USE_EXISTING_COGNITO=true
• EXISTING_COGNITO_USER_POOL_ID=

the script overrides this configuration and creates a new Cognito User Pool instead of reusing the existing one.

Steps to Reproduce

1. First deployment (no env variables):

   • Run deployment without USE_EXISTING_COGNITO
   • A new Cognito User Pool is created successfully

2. Second deployment (reuse existing pool):

   • Set:
     export USE_EXISTING_COGNITO=true
     export EXISTING_COGNITO_USER_POOL_ID=
   • Run deployment again

3. Observe behavior:

   • Script resets USE_EXISTING_COGNITO=false
   • A new Cognito User Pool is created again

Expected Behavior

When both variables are provided:

• USE_EXISTING_COGNITO=true
• EXISTING_COGNITO_USER_POOL_ID is valid

the script should:

• Reuse the existing Cognito User Pool
• Skip auto-detection and creation logic

Current Behavior

The script checks stack existence and overrides:

• USE_EXISTING_COGNITO=false

even when the user has explicitly provided configuration.

Impact

• Prevents reuse of existing Cognito resources
• Leads to unnecessary creation of multiple User Pools
• Breaks expected idempotent deployment behavior

Proposed Fix

Update logic to:

• Respect user-provided configuration when both variables are set
• Avoid overriding USE_EXISTING_COGNITO in that case
• Only run auto-detection when configuration is not provided