We use a sharedFlow within the "post-proxy flowhook". The last time we had to make a change, it did not go well because when the sharedFlow was deployed to the "production" environment, one message processor did not deploy it successfully. We updated the flowhook and started having a lot of intermittent failures from proxies that tried to access the flowhook on the 'bad' MP.
This type of deployment failure had not happened in our lower environments, but it made us realize that our testing process should be better.
Right now I'm thinking of doing it like this:
1. Always create an entirely new version of the sharedFlow when it changes, i.e., create a new sharedFlow with a new name where 'sharedFlow-v1' becomes 'sharedFlow-v2'.
2. Use a no-target proxy as a 'sharedflow-tester' and set it up to invoke the new version of the sharedFlow.
3. Deploy the new sharedFlow and deploy the 'sharedflow-tester' proxy to each environment, so then we can invoke 'sharedflow-tester'.
4. If tests to 'sharedflow-tester' are successful, update the flowhook to use the new sharedFlow.
I suspect I'm not alone in trying to test sharedFlows efficiently, so I'm curious; Does anyone have tips or any "best practice" recommendations?
LinkedIn
Twitter
