Skip to content

Introduction

Welcome onboard of the Capella Collaboration Manager. This platform helps you co-work on architectures using tools like Capella, Papyrus, pure::variants and Jupyter in combination with py-capellambse.

This platform wraps the modeling tools to create an environment where users can can work directly in the project context without having to install or configure the tool. It also takes care of the housekeeping of the modeling projects themselves. And that's not all: We also offer project access management and since the site is completely responsive it is also usable on smartphones.

Automations and Model-derived Products

Git is quite in the middle of the modeling lifecycle. We also use its automation means (CI/CD) to automate a number of housekeeping activities:

  • Automated model-modifications: This includes range of services like human-friendly element ID assignment, change control and versioning of elements, hyperlinked object title update in descriptions and maintenance of model-derived requirements (req-bot).
  • Derived product generation and distribution: generation and publication of model-derived documents and other artifacts (like software interfaces, configurations); caching of diagrams for faster display in linked pages and web-viewer; computation of model metrics for modeling progress dashboards; spell-checking; synchronization with tools like Simulink, Polarion, Codebeamer, Confluence or even Grafana.

At this moment Collaboration Manager doesn't provide you with a self-service to configure any of these automations, however since now you know these are possible - you may get in touch with your Systems Engineering Toolchain contact to get these configured.

Now that you have some understanding of the modeling setup and ways of working we can look into how you can actually work in this environment.

Working with Capella in the Collaboration Manager

Capella is not a web native tool, however it runs on Linux - so we can use open-source technologies to deliver it to you via browser. We constructed a Docker container with Capella and all the plugins you may need, added Xrdp to allow remote connection and Apache Guacamole to stream that connection to your browser. We call instances of those containers Sessions. To make this platform scalable and allow people co-work comfortably we use Kubernetes to run the Session containers - which in Kubernetes terms are called Pods. Kubernetes is able to scale the cluster to accommodate the active sessions, making it run at reasonable costs in both public and private cloud environments (but of course one could run it in a VM).

There are currently 2 session types supported:

  • Read-only session - in this case Collaboration Manager gets the latest (or user-selected) model version from git and places that into a read-only workspace within Capella. You can "play" with that model and even make changes, however these changes will not be saved and so will do no harm (for instance to agreed / approved contents). When the session is closed the contents of the workspace is gone.
  • Persistent workspace - in this mode a user-specific persistent volume is mounted to the Session pod and linked to Capella as the workspace. This enables you to work on projects locally, via git or TeamForCapella as persistent workspace keeps "state" even after the session is closed.

We do currently support two different working modes: TeamForCapella and Git. If you want to get more information about it, we have prepared a comparison here: Ways of co-working on a Capella project

User Roles

There are 3 roles you can have in a project context:

  • User - read-only - you can view model snapshots (latest model, any specific release, branch or commit) from git. You may edit the model however your changes will not be saved. (Makes it also useful for training exercises.)
  • User - read/write - you start a Persistent Workspace Session. Your user account is allowed to clone and commit to a git project — if the project co-working model is git-only — or allowed to connect to a remote repository in a TeamForCapella-based project. Also in this mode you may have many co-working projects open at the same time, given that you have a role in those projects that allows this kind of access.
  • Model manager - can do same as both users above but also can invite or remove users from the managed projects and control their access rights.

You may also learn more about the roles model here.

Next Steps

If you like to start a new project and work on it via Collab Manager - please follow this guide: project onboarding guideline to prepare your project. If you want to use the TeamForCapella workflow, please contact your administrator. Linking models to TeamForCapella repositories can only be done by administrators.

To work with the application, you need access to a project. If you don't have it yet or can't find the project you need see how do I get access to a project.

At this point you may want to continue to one of the detailed getting-started sections:

Missing Information / Need Support

Important

The below is only valid for content-agnostic issues, do not share any proprietary / project related data or real content in reported issues or screenshots.

If you don't find answer to your question on this documentation site please consider opening an issue on GitHub or extending the documentation with your own contribution via a pull request.