Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update MLX setup instructions for KF 1.5 #346

Merged
merged 2 commits into from
Jun 24, 2022
Merged

Update MLX setup instructions for KF 1.5 #346

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1d9a628
Update MLX setup instructions for KF 1.5
ckadner Jun 23, 2022
87ceeac
Update supported Kubernetes version
ckadner Jun 23, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions docs/install-mlx-on-kind.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -63,7 +63,7 @@ kubectl get pods --all-namespaces
## Deploy MLX (Single-User)

```Bash
git clone https://github.com/IBM/manifests -b v1.4.0-mlx
git clone https://github.com/IBM/manifests -b v1.5-branch
cd manifests

# run the below command two times if the CRDs take too long to provision
Expand All @@ -80,7 +80,8 @@ done
kubectl get pods --all-namespaces

# make the MLX UI available to your local browser on http://localhost:3000/
kubectl port-forward -n istio-system svc/istio-ingressgateway 3000:80 &
kubectl wait -n kubeflow --for=condition=ready pod -l service=mlx-ui && \
kubectl port-forward -n istio-system svc/istio-ingressgateway 3000:80 &
```

Now paste the URL http://localhost:3000/login into your browser and proceed to
Expand Down
47 changes: 36 additions & 11 deletions docs/mlx-setup.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,8 +1,9 @@
# Deploy MLX on an existing Kubernetes cluster

## Prerequisites

* An existing Kubernetes cluster:
- **Min version: 1.17**
- **Min version: 1.21**
- **Max version: 1.21**
* The recommended minimum capacity requirement for MLX are:
- **CPUs**: 8 Cores
Expand All @@ -19,9 +20,9 @@
mv kustomize /usr/local/bin
```

## Deploy
## Deployment

To deploy the MLX single-user mode on an existing Kubernetes Cluster, clone the MLX manifests and deploy it with Kustomize.
To deploy the MLX **single-user** mode on an existing Kubernetes Cluster, clone the MLX manifests and deploy it with Kustomize.

This MLX deployment includes:
- [Istio](https://istio.io/)
Expand All @@ -30,11 +31,33 @@ This MLX deployment includes:
- [Datashim](https://datashim-io.github.io/datashim/) to provide access to S3 and NFS Datasets within pods
- MLX API and UI

**Note:** Before deploying MLX on OpenShift on Fyre, you need to install a persistent
storage provider like [Portworx](https://docs.portworx.com/install-portworx/openshift/)
and set it as the `default` storage class. This is only required to support mounting
Persistent Volumes for Datasets using Datashim.

To deploy MLX on your Kubernetes or OpenShift cluster, first set the `MLX_DEPLOYMENT_TYPE`
environment variable based on your Kubernetes service provider by uncommenting
one of the `export` commands below:

```Shell
# export MLX_DEPLOYMENT_TYPE=mlx-single-ibmcloud # IBM Cloud - Kubernetes
# export MLX_DEPLOYMENT_TYPE=mlx-single-ibmcloud-openshift # IBM Cloud - OpenShift
# export MLX_DEPLOYMENT_TYPE=mlx-single-fyre-openshift # IBM Fyre - OpenShift

echo "MLX deployment type: ${MLX_DEPLOYMENT_TYPE:-"UNDEFINED"}"
```

Then clone the [IBM/manifest](https://github.com/IBM/manifests/tree/v1.5-branch)
repository and apply the manifests to your cluster:

```shell
git clone https://github.com/machine-learning-exchange/manifests -b mlx-single-user
cd manifests
# run the below command two times if the CRDs take too long to provision.
while ! kustomize build example | kubectl apply -f -; do echo "Retrying to apply resources"; sleep 10; done
# clone the manifest repo
git clone https://github.com/IBM/manifests -b v1.5-branch && cd manifests

# run the following command twice if the CRDs take too long to provision
while ! kustomize build ${MLX_DEPLOYMENT_TYPE} | \
kubectl apply -f -; do echo "Retrying to apply resources"; sleep 10; done
```

Then access the MLX web page on http://<cluster_node_ip>:30380/mlx/
Expand All @@ -44,14 +67,16 @@ This MLX deployment doesn't include or support:
- Multi-user mode
- Istio mutual TLS

To get these features, please install the additional plugins by following the instructions for [MLX deployment an existing Kubeflow cluster](/docs/install-mlx-on-kubeflow.md#deploy-mlx-on-an-existing-kubeflow-cluster).
To get those features, please install the additional plugins by following the instructions for
[MLX deployment an existing Kubeflow cluster](/docs/install-mlx-on-kubeflow.md#deploy-mlx-on-an-existing-kubeflow-cluster).

## Delete the MLX deployment

## Deleting the MLX Deployment

To delete this MLX deployment, run the following commands in the same manifests folder.

```
kustomize build example | kubectl delete -f -
```Shell
kustomize build ${MLX_DEPLOYMENT_TYPE} | kubectl delete -f -
```

## Troubleshooting
Expand Down
3 changes: 2 additions & 1 deletion docs/mlx-workshop.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -202,7 +202,8 @@ For MLX on KIND we need to make the port available on localhost by running the
following command:

```Bash
kubectl port-forward svc/codenet-language-classification 5000:5000
kubectl wait --for=condition=ready pod -l app=codenet-language-classification && \
kubectl port-forward svc/codenet-language-classification 5000:5000
```

Now we should be able to access the UI page for the deployed model by pasting
Expand Down