We will have dev, testing, production R Apps. Will we need a separate DEV/TEST environment for each product? How do can we utilize these products to satisfy our approach on dev, testing and production on R Applications?
This is an interesting question - to help you more thoroughly, we likely need to hear more details, particularly around what the requirements currently are for a dev/test/prod framework at your organization:
- Are there defined requirements that you'll need to demonstrate, or are you looking for guidance around what best practices could be?
- If you aren't comfortable sharing that level of detailed information on this forum, please reach out to your Customer Success representative.
This resource on staging servers and how to use them in a dev/test/prod set up might be a good place to start the conversation.
Should I use a staging license with my Dev/Test/Prod architecture?
You can use a staging server to test your applications before deployment into production. If you want to set up a dev/test/prod framework for deploying content like Shiny apps, we recommend using the RStudio IDE for development, an RStudio Connect staging server for test, and an RStudio Connect server (or cluster) for production.
I develop applications locally.
Merge the local branches to staging branch. Test the staging app online that is restricted to some users. Test users are satisfied with the application, then we will merge the staging branch to the production branch. That's me but I want to hear from RStudio for best practices.
As for the staging server, I see RStudio Connect (Staging) and RStudio Connect (Production) does that mean that I need to have 2 RStudio Connect license?
Thank you for the fast response. Really happy that the community page is helpful and responsive!
Have you done an evaluation of the RStudio Connect product yet? To get a sense of how the App development and deployment workflow will fit into your current workflow, I would recommend getting the 45 day free evaluation. You may find that the controls available to you in RStudio Connect (e.g. controls around application visibility and access) might negate the need for a separate, second staging server if your goals are limited to user satisfaction testing.
Staging servers are great to have, and the reference I linked to above enumerates some of the reasons why we see them as an important part of analytic infrastructure. The licenses for staging servers work a little differently. The sales team
email@example.com would be glad to have a conversation with you about all the options and what kind of configuration plan might ultimately serve your needs best.