Overall Architecture

Principles

  • Containers that operate like serverless
  • Link to hosted cloud resources rather than trying to build alternative interfaces, where possible
  • None of your customer data leaves your own cloud projects and network boundaries (by Coherence's actions, at least!)

Roadmap Goals

  • Make compliance easy - if you use Coherence, eventually you should be able to get SOC2/ISO27001/HITRUST for the SDLC-related parts of your org as easily as AWS/GCP can provide that certification for your data center today. More than compliance, our focus on best practices means you're constantly getting the benefits of our focus to not only meet these standards, but to exceed them and get real-world benefits for your users instead of focusing on box-checking.

Infra Flavors

Coherence is not focused on multi-cloud deployments of the same application environments for the purposes of reliability, scalability, or cost-savings. Dynamic cross-cloud workload management is its own space and not likely to be a feature we offer in the medium term. This means that migrating state (DB/cache/message queue) across clouds is unlikely to be our purview.

We do intend to be able to deploy the same environment to different infrastructure configurations based on the needs of your application. This will be useful from a business continuity perspective, as well as offering the same coherent experience with our abstractions and best practices across different deployment targets (maybe for example two different apps within a company that deploy to different infra flavors in the same cloud or to different clouds).

At launch (April 2022), we intend to support one infra flavor - Google Cloud Run. See Google Cloud Run for details.

Key Features of All Flavors

  • Support for multiple apps deployed to different load balancers and cloud projects (see Repo Structure). Support for multiple services per repo (e.g. frontend/backend - multiple of either type) routing behind the same IP to different paths
  • Separate production project/environment
  • Per user/branch dev environments (VSCode based, can run VIM or emacs, JetBrains coming on roadmap). All service containers and dependencies running in one kubernetes pod
  • Managed no-code gitops - for configured services, we auto-manage your infra based on pushes to configured branches. Branch environments that auto-update on push. These are production-equivalent test environments in a separate shared project for all review environments
  • Managed CI YAML generation, including database/cache dependencies for per-service unit tests, running in your own CI platform. Ability to run integration tests on a deployed branch environment coming soon

What’s Next
Did this page help you?