docs: hello world deploy example #2102
Conversation
atchernych
changed the title
WalkthroughThis update revises documentation for example deployments, adds Kubernetes deployment instructions and a manifest for the Hello World example, and improves clarity on image building and usage. The changes include new example links, updated resource references, and detailed deployment steps for running the Hello World example in a Kubernetes environment. Changes
Sequence Diagram(s)sequenceDiagram
participant User
participant Docs
participant Kubernetes
participant Frontend
participant HelloWorldWorker
User->>Docs: Read deployment instructions
User->>Kubernetes: Apply hello_world.yaml manifest
Kubernetes->>Frontend: Deploy Frontend service
Kubernetes->>HelloWorldWorker: Deploy HelloWorldWorker service
Frontend-->>Kubernetes: Report liveness/readiness
HelloWorldWorker-->>Kubernetes: Report liveness/readiness
User->>Kubernetes: (Optional) Delete deployment
Estimated code review effort🎯 2 (Simple) | ⏱️ ~8 minutes Possibly related PRs
Poem
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
Documentation and Community
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (5)
examples/runtime/hello_world/README.md (1)
104-118: Minor polish for the new K8s deployment section
- The relative link
../../../docs/guides/dynamo_deploy/quickstart.mdwill break when this README is rendered on GitHub (it escapes the repo root). Consider an absolute repo-relative link (docs/guides/dynamo_deploy/quickstart.md) or a full URL.cd dynamois redundant if the user is already in the repository root; it also breaks paths when the repo is cloned into a differently named folder.- Worth adding a note that the namespace must exist:
kubectl create namespace ${NAMESPACE}.No functional issues, just usability tweaks.
docs/examples/README.md (2)
7-10: Typo & wording
owm→own-You could build your owm image using +You could build your own image usingAlso, “prior to applying” can be shortened to “before applying” for brevity.
43-45: Make the CR-YAML hint actionableYou mention that many examples ship a CR under
deploy/, but only link to the vLLM YAML. It would be helpful to list (or at least link) the newly-added Hello-World YAML here for discoverability.-Use [VLLM YAML](../../components/backends/vllm/deploy/agg.yaml) for an example. +See the [VLLM YAML](../../components/backends/vllm/deploy/agg.yaml) or the new +[Hello-World YAML](../../examples/runtime/hello_world/deploy/hello_world.yaml) +for concrete examples.examples/runtime/hello_world/deploy/hello_world.yaml (2)
41-47: Image reference is private – consider a public example
gitlab-master.nvidia.com:5005/...is not accessible to external users reading the OSS docs.
Pointing to a publicnvcr.io/...image (or using a placeholder<your-image>) would make the manifest immediately runnable.
48-67: Worker readiness probe may never succeedThe probe looks for
"Serving endpoint"inside/tmp/hello_world.log, but the container command appends (tee) to that file only after Python starts. If the log message format changes or buffering delays the write, the pod will stayNotReady.Consider either:
• Switching to an HTTP readiness probe (mirroring the Frontend), or
• Using a deterministic file touch/echo at the end of worker startup.
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (4)
docs/examples/README.md(4 hunks)docs/index.rst(1 hunks)examples/runtime/hello_world/README.md(1 hunks)examples/runtime/hello_world/deploy/hello_world.yaml(1 hunks)
🧰 Additional context used
🧠 Learnings (3)
examples/runtime/hello_world/README.md (3)
Learnt from: julienmancuso
PR: #2012
File: deploy/cloud/helm/crds/templates/nvidia.com_dynamocomponentdeployments.yaml:92-98
Timestamp: 2025-07-18T16:04:31.771Z
Learning: CRD schemas in files like deploy/cloud/helm/crds/templates/*.yaml are auto-generated from Kubernetes library upgrades and should not be manually modified as changes would be overwritten during regeneration.
Learnt from: julienmancuso
PR: #2012
File: deploy/cloud/helm/crds/templates/nvidia.com_dynamocomponentdeployments.yaml:1178-1180
Timestamp: 2025-07-18T16:05:05.534Z
Learning: The stopSignal field under lifecycle in DynamoComponentDeployment CRDs is autogenerated due to Kubernetes library upgrades (k8s.io/api and k8s.io/apimachinery from v0.32.3 to v0.33.1), not a manual design decision by the user.
Learnt from: nnshah1
PR: #1444
File: tests/fault_tolerance/utils/metrics.py:30-32
Timestamp: 2025-07-01T13:55:03.940Z
Learning: The @dynamo_worker() decorator in the dynamo codebase returns a wrapper that automatically injects the runtime parameter before calling the wrapped function. This means callers only need to provide the non-runtime parameters, while the decorator handles injecting the runtime argument automatically. For example, a function with signature async def get_metrics(runtime, log_dir) decorated with @dynamo_worker() can be called as get_metrics(log_dir) because the decorator wrapper injects the runtime parameter.
docs/examples/README.md (1)
Learnt from: PeaBrane
PR: #1409
File: examples/router_standalone/worker.py:171-186
Timestamp: 2025-06-08T08:30:45.126Z
Learning: Example code in the examples/ directory may intentionally use hard-coded values or simplified implementations that wouldn't be appropriate for production code, but are acceptable for demonstration and testing purposes.
examples/runtime/hello_world/deploy/hello_world.yaml (2)
Learnt from: julienmancuso
PR: #2012
File: deploy/cloud/helm/crds/templates/nvidia.com_dynamocomponentdeployments.yaml:1178-1180
Timestamp: 2025-07-18T16:05:05.534Z
Learning: The stopSignal field under lifecycle in DynamoComponentDeployment CRDs is autogenerated due to Kubernetes library upgrades (k8s.io/api and k8s.io/apimachinery from v0.32.3 to v0.33.1), not a manual design decision by the user.
Learnt from: julienmancuso
PR: #2012
File: deploy/cloud/helm/crds/templates/nvidia.com_dynamographdeployments.yaml:1233-1235
Timestamp: 2025-07-18T16:04:47.465Z
Learning: The stopSignal field in Kubernetes CRDs like DynamoGraphDeployment and DynamoComponentDeployment is autogenerated by controller-gen when upgrading Kubernetes library versions, and represents expected upstream API changes rather than manual code that needs custom validation.
🪛 LanguageTool
docs/examples/README.md
[style] ~50-~50: ‘prior to’ might be wordy. Consider a shorter alternative.
Context: ...date the image location in your CR file prior to applying. You could build your owm imag...
(EN_WORDINESS_PREMIUM_PRIOR_TO)
[grammar] ~51-~51: Ensure spelling is correct
Context: ...prior to applying. You could build your owm image using ```bash ./container/build....
(QB_NEW_EN_ORTHOGRAPHY_ERROR_IDS_1)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
- GitHub Check: Mirror Repository to GitLab
🔇 Additional comments (3)
docs/examples/README.md (1)
49-63: Flag large CPU/GPU figures in documentation sample10 vCPU, 20 GiB RAM and 1 GPU for “Hello World” may scare newcomers and exhaust small clusters. Consider using modest defaults (e.g., 0.5 CPU, 512 Mi, 0 GPU) and documenting how to raise them for production.
examples/runtime/hello_world/deploy/hello_world.yaml (2)
72-79: Resource requests are excessive for a demo10 CPU / 20 GiB / 1 GPU for a greeting service is over-provisioned and blocks adoption on Minikube-like clusters. Recommend lowering defaults and documenting how to upscale.
32-39: Use the standard GPU resource keyKubernetes expects vendor-qualified extended resources, e.g.
nvidia.com/gpu.
Using a plaingpu:key will be ignored by the scheduler unless a custom resource-quota controller is in place.- gpu: "1" + nvidia.com/gpu: "1"⛔ Skipped due to learnings
Learnt from: nnshah1 PR: ai-dynamo/dynamo#1444 File: tests/fault_tolerance/configs/agg_tp_1_dp_8.yaml:31-38 Timestamp: 2025-07-01T15:33:53.262Z Learning: In fault tolerance test configurations, the `resources` section under `ServiceArgs` specifies resources per individual worker, not total resources for all workers. So `workers: 8` with `gpu: '1'` means 8 workers × 1 GPU each = 8 GPUs total.
Signed-off-by: atchernych <[email protected]>
Signed-off-by: atchernych <[email protected]>
Signed-off-by: atchernych <[email protected]>
Overview:
DYN-704
Details:
Where should the reviewer start?
Related Issues: (use one of the action keywords Closes / Fixes / Resolves / Relates to)
Summary by CodeRabbit
Documentation
New Features