Add documentation for transformer collocation with runtime #464
+263
−10
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
❌ Deploy Preview for elastic-nobel-0aef7a failed.
|
Enhance transformer documentation Signed-off-by: Sivanantham Chinnaiyan <[email protected]>
sivanantha321
force-pushed
the
collocation-runtime-docs
branch
from
59c24d5 to
79654fc
Compare
spolti
reviewed
Apr 11, 2025
|
|
||
| Since, the predictor and the transformer are in the same pod, they need to listen on different ports to avoid conflict. `Transformer` is configured to listen on port 8080 (REST) and 8081 (GRPC) | ||
| while, `Predictor` listens on port 8085 (REST). `Transformer` calls `Predictor` on port 8085 via local socket. | ||
| Deploy the `Inferenceservice` using the below command. | ||
|
|
||
| ```bash | ||
| Note that, readiness probe is specified in the transformer container. This due to the limitation of Knative. You can provide `--enable_predictor_health_check` argument to allow the transformer container to check the predictor health as well. This will make sure that both the containers are healthy before the isvc is marked as ready. |
There was a problem hiding this comment.
Suggested change
| Note that, readiness probe is specified in the transformer container. This due to the limitation of Knative. You can provide `--enable_predictor_health_check` argument to allow the transformer container to check the predictor health as well. This will make sure that both the containers are healthy before the isvc is marked as ready. | |
| Note that, readiness probe is specified in the transformer container. This is due to the limitation of Knative. You can provide `--enable_predictor_health_check` argument to allow the transformer container to check the predictor health as well. This will make sure that both the containers are healthy before the isvc is marked as ready. |
spolti
reviewed
Apr 11, 2025
| @@ -52,13 +55,20 @@ spec: | |||
| image: kserve/image-transformer:latest | |||
| args: | |||
| - --model_name=mnist | |||
| - --protocol=v1 # protocol of the predictor; used for converting the input to specific protocol supported by the predictor | |||
| - --predictor_protocol=v1 | |||
There was a problem hiding this comment.
May we keep the comment or use v2?
spolti
reviewed
Apr 11, 2025
| ports: | ||
| - containerPort: 8080 | ||
| protocol: TCP | ||
| readinessProbe: |
There was a problem hiding this comment.
What about adding liveness too?
spolti
reviewed
Apr 11, 2025
| ### Deploy the InferenceService | ||
|
|
||
| Since, the predictor and the transformer are in the same pod, they need to listen on different ports to avoid conflict. `Transformer` is configured to listen on port 8080 (REST) and 8081 (GRPC) | ||
| while, `Predictor` listens on port 8085 (REST). `Transformer` calls `Predictor` on port 8085 via local socket. |
There was a problem hiding this comment.
Maybe also mention which is the default port for gRPC for predictors.
spolti
reviewed
Apr 11, 2025
|
|
||
| Since, the predictor and the transformer are in the same pod, they need to listen on different ports to avoid conflict. `Transformer` is configured to listen on port 8080 (REST) and 8081 (GRPC) | ||
| while, `Predictor` listens on port 8085 (REST). `Transformer` calls `Predictor` on port 8085 via local socket. | ||
| Deploy the `Inferenceservice` using the below command. |
There was a problem hiding this comment.
Suggested change
| Deploy the `Inferenceservice` using the below command. | |
| Deploy the `Inferenceservice` using the following command: |
spolti
reviewed
Apr 11, 2025
| while, `Predictor` listens on port 8085 (REST). `Transformer` calls `Predictor` on port 8085 via local socket. | ||
| Deploy the `Inferenceservice` using the below command. | ||
|
|
||
| Note that, readiness probe is specified in the transformer container. This due to the limitation of Knative. You can provide `--enable_predictor_health_check` argument to allow the transformer container to check the predictor health as well. This will make sure that both the containers are healthy before the isvc is marked as ready. |
spolti
reviewed
Apr 11, 2025
|
|
||
| !!! note | ||
| Do not specify ports for predictor in the serving runtime for Serverless deployment. This is not supported by knative. | ||
| For more info see, [knative discussion on multiple ports](https://github.com/knative/serving/issues/8471). |
There was a problem hiding this comment.
Suggested change
| For more info see, [knative discussion on multiple ports](https://github.com/knative/serving/issues/8471). | |
| For more information, please take a look at [knative discussion on multiple ports](https://github.com/knative/serving/issues/8471). |
spolti
reviewed
Apr 11, 2025
| - name: "TS_SERVICE_ENVELOPE" | ||
| value: "{% raw %}{{.Labels.serviceEnvelope}}{% endraw %}" | ||
| securityContext: | ||
| runAsUser: 1000 # User ID is not defined in the Dockerfile, so we need to set it here to run as non-root |
There was a problem hiding this comment.
I think k8s adds it by default, ideally, we need to run a random uid, no?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Enhance transformer documentation
"Fixes kserve/kserve#4343"
Proposed Changes