## Problem Some of the larger organizations using GitLab, whether they are on SaaS or Self-managed, may have existing infrastructure that can run remote development environments. Using this existing infrastructure would have security and efficiency benefits without requiring potentially costly GitLab-hosted infrastructure. ## Purpose Since we can't "simply" [use the first-party Remote Container extension](https://gitlab.com/gitlab-org/gitlab/-/issues/360866), we need to investigate how we can connect the Web IDE to an existing container provisioned on a customer's existing infrastructure and get real-time feedback in the terminal. ## Scope The goal of this effort is to: 1. Define the commonalities across customers' existing infrastructures and their requirements for connectivity (likely SSH) 2. Investigate technical approaches to connecting the Web IDE to a container running a development environment. 3. Create an epic that outlines an iteration plan for an MVC that makes it possible to connect to at least a narrow subset of development environments defined in step 1. ## Questions 1. We can't use SSH. What are the viable alternatives for establishing a secure websocket connection? 2. Are we running into any limitations in functionality with the Web IDE by not having it running in the container itself? For example, code completion, linting, and language server type features - are we going to get that to work with this approach? 3. How will these environments be provisioned, managed, created, destroyed, etc? To what extent do we need to have UI around this in our MVC?