docs: add k3s use case guide - #311
Conversation
|
Preview deployment for your docs. Learn more about Mintlify Previews.
|
| icon: "dharmachakra" | ||
| --- | ||
|
|
||
| The `e2b/k3s` template turns an E2B Sandbox into a preconfigured, single-node [k3s](https://k3s.io) cluster. `kubectl`, cluster DNS, networking, storage, and the standard k3s components are configured when the sandbox starts. |
There was a problem hiding this comment.
I'm not sure if standard k3s componets is a real thing. Quick fix is standard Kubernetes components or standard k8s components.
There was a problem hiding this comment.
Good catch. My depth with k3s/Kubernetes is limited, so I may be using the terminology too loosely here. Would you prefer “the components bundled with k3s” or “the default k3s components”? I want the wording to reflect the implementation accurately rather than imply there’s a canonical standard set.
There was a problem hiding this comment.
No need to be too precise here IMO, but since we are referring to:
cluster DNS, networking, storage, ...
which are more generically Kubernetes concepts and not k3s, I'd say default k8s components.
k3s is just more lightweight, easier to manage, distribution of Kubernetes. Similarly to how Ubuntu is a distribution of Linux.
There was a problem hiding this comment.
use default k8s components, addressed at 099fcca
|
|
||
| The `e2b/k3s` template turns an E2B Sandbox into a preconfigured, single-node [k3s](https://k3s.io) cluster. `kubectl`, cluster DNS, networking, storage, and the standard k3s components are configured when the sandbox starts. | ||
|
|
||
| Use it when an agent or CI job needs a real Kubernetes API without provisioning or sharing a permanent cluster. |
There was a problem hiding this comment.
Would you prefer “an isolated Kubernetes API” or “a single-node Kubernetes cluster” here?
There was a problem hiding this comment.
a single-node Kubernetes cluster
There was a problem hiding this comment.
use single-node Kubernetes cluster, addressed at 099fcca
| `kubectl` is already configured. On a fresh or resumed sandbox, wait for the node and cluster DNS objects to appear and become ready before deploying: | ||
|
|
||
| ```bash | ||
| timeout 120s sh -c 'until kubectl get node -o name | grep -q .; do sleep 1; done' |
There was a problem hiding this comment.
Good for now, in future we may want to absorb these "wait until ready" in our bootstrap phase. So users will get a ready-to-use cluster.
There was a problem hiding this comment.
I’ll keep them in the docs for now, then remove them once the template bootstrap guarantees the complete cluster is ready. Happy to track that as a follow-up.
Summary
/docs/use-cases/k3sguide for the officiale2b/k3stemplateSandboxAPI andcommands.runkubectl port-forwardandgetHostDependencies
e2b/k3sis published for public use.Validation
mintlify validatemintlify broken-links9121d260) under a temporary private tagThe docs explicitly poll for the Node and CoreDNS objects before waiting for their Ready conditions because PR #49 currently gates sandbox creation on Kubernetes API readiness only.