[Choreo Connect 1.2.0] Rolling Update without downtime with Choreo

[DaniduWijendra]

Description

When configuring Choreo Connect with both router and enforcer that run on the same Pod and adapter on another pod, we observed error 500 responses for API invocations with the error code 102500 ("Router couldn't connect to the Enforcer" as it is written in the documentation [1] ) during a rolling update.

This issue happened only with API invocation during the rolling update. When the API was reachable via Choreo, we did not observe this behavior until the next update.

[1] https://apim.docs.wso2.com/en/4.1.0/deploy-and-publish/deploy-on-gateway/choreo-connect/troubleshooting/error-handling/#error-codes

Steps to Reproduce

• Deploy choreo connect on Kubernetes environment with APIM.

kubectl apply -f choreo-connect-with-apim/choreo-connect/

• Deploy the APIM pod

kubectl apply -f choreo-connect-with-apim/apim/

• Deploy an API on the publisher portal and invoke it via a load test
• Then made some changes on the choreo-connect-deployment.yaml file and perform a rolling update.
• Command run to apply the changes

kubectl apply -f choreo-connect-with-apim/choreo-connect/

In that case, you can see the following error happens with the 102500 error code.

Version

1.2.0

Environment Details (with versions)

No response

Relevant Log Output

No response

Related Issues

No response

Suggested Labels

No response