Salesforce Certified Heroku Architecture Designer — Question 38

A client wants to migrate their on-premise application to Heroku. The application maintains a local, in-memory cache for its data store to improve performance.
What should an Architect advise the client about running the application on Heroku?

Answer options

• A. The application should declare its in-memory cache as a separate process type to preserve it across deploys.
• B. All of the application's web dynos will share a single in-memory cache.
• C. The application should replicate its in-memory cache to a local disk to preserve it across deploys.
• D. The application's in-memory cache might be cleared at any time.

Correct answer: A

Explanation

The correct answer is A because declaring the in-memory cache as a separate process type allows it to persist across deploys, which is crucial for maintaining performance. Option B is incorrect as each dyno in Heroku runs in isolation and does not share in-memory data. Option C is also wrong because Heroku's ephemeral filesystem means that local disk storage does not guarantee persistence during deploys. Option D highlights a risk, but it doesn't provide a solution for preserving the cache.