API / Apollo Gateway Interview questions
What is the difference between self-hosted Apollo Router and the GraphOS Router (cloud)?
Both run the same Router technology, but who operates it — and how much infrastructure control you keep — differs.
| Self-hosted Apollo Router | GraphOS Router (cloud) |
| You run the router binary/container yourself, on your own infra | Apollo runs and scales the router for you as a managed service |
| Connects to GraphOS mainly for supergraph delivery (Uplink) and metrics reporting | Fully managed end to end, including scaling and availability |
| Full control over deployment region, scaling policy, and networking | Simpler operations, less direct infrastructure control |
| You own capacity planning and on-call for the router itself | Apollo owns router-level operations; you focus on subgraphs |
The choice generally comes down to how much operational ownership a platform team wants over the routing layer versus how much they'd rather hand off — the Federation behavior itself (composition, entity resolution, query planning) is the same either way.
In the self-hosted model, what does the router primarily use GraphOS for?
What stays the same regardless of whether you choose self-hosted or GraphOS Router (cloud)?
More Related questions...
