Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
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
2 changes: 1 addition & 1 deletion Makefile
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,7 @@ endif
OPERATOR_SDK_VERSION ?= v1.41.1

# NOTE: This is a source of truth for some of the workflows. Don't change the format!
TAG ?= 0.3.0-rc1
TAG ?= 0.3.0

# Image URL to use all building/pushing image targets
IMG ?= ghcr.io/intel/intel-gpu-base-operator:$(TAG)
Expand Down
35 changes: 20 additions & 15 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -92,13 +92,9 @@ Helm deployment is split into two charts: operator and policy. The reason for th

The basic installation is as follows:
```
kubectl create ns intel-gpu-operator
# Required by DRA's admin access
kubectl label ns intel-gpu-operator resource.kubernetes.io/admin-access=true

helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-operator \
helm install --create-namespace --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-operator \
oci://ghcr.io/intel/intel-gpu-base-operator-chart --wait
helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-policy \
helm install --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-policy \
oci://ghcr.io/intel/intel-gpu-base-operator-policy-chart --set resourceRegistration=dra
```

Expand All @@ -107,20 +103,20 @@ This installs the operator and a DRA-enabled deployment with Intel XPU Manager.
#### Example: DRA without NFD

```
helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-operator \
helm install --create-namespace --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-operator \
oci://ghcr.io/intel/intel-gpu-base-operator-chart --wait
helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-policy \
helm install --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-policy \
oci://ghcr.io/intel/intel-gpu-base-operator-policy-chart --set resourceRegistration=dra
Comment thread
tkatila marked this conversation as resolved.
```

#### Example: DRA with NFD and Kueue

```
helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-operator \
helm install --create-namespace --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-operator \
oci://ghcr.io/intel/intel-gpu-base-operator-chart --wait \
--set nfd.install=true \
--set kueue.install=true
helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-policy \
helm install --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-policy \
oci://ghcr.io/intel/intel-gpu-base-operator-policy-chart \
--set resourceRegistration=dra \
--set useNFDLabeling=true \
Expand All @@ -130,24 +126,33 @@ helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-policy \
#### Example: Device Plugin with NFD

```
helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-operator \
helm install --create-namespace --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-operator \
oci://ghcr.io/intel/intel-gpu-base-operator-chart --wait \
--set nfd.install=true
helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-policy \
helm install --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-policy \
oci://ghcr.io/intel/intel-gpu-base-operator-policy-chart \
--set resourceRegistration=dp \
--set useNFDLabeling=true
```

Uninstalling the charts:
```
helm uninstall --namespace "intel-gpu-operator" gpu-policy --wait
helm uninstall --namespace "intel-gpu-operator" gpu-operator
kubectl delete ns intel-gpu-operator
helm uninstall --namespace "intel-gpu-base-operator" gpu-policy --wait
helm uninstall --namespace "intel-gpu-base-operator" gpu-operator
```

See more details for the charts in the [operator](charts/gpu-base-operator/README.md) and [policy](charts/gpu-base-operator-policy/README.md) READMEs.

#### Installation to an existing namespace

Installation to a pre-existing namespace needs to set `createNamespace=false` chart variable and drop the `--create-namespace` argument to prevent chart from trying to create the namespace.

Also note that DRA requires the install namespace to contain `resource.kubernetes.io/admin-access=true` label for the monitoring to function. If one installs the operator to an existing namespace, the namespace has to be labeled:

```
kubectl label ns <target namespace> resource.kubernetes.io/admin-access=true
```

### Custom Resource (CR) fields

CR fields control how the operator configures the cluster. See the [full struct](api/v1alpha1/clusterpolicy_types.go) for all options.
Expand Down
4 changes: 2 additions & 2 deletions charts/gpu-base-operator-policy/Chart.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -2,5 +2,5 @@ apiVersion: v2
name: intel-gpu-base-operator-policy-chart
description: A Helm chart for Intel GPU Base Operator deployments
type: application
version: 0.3.0-rc1
appVersion: "0.3.0-rc1"
version: 0.3.0
appVersion: "0.3.0"
6 changes: 3 additions & 3 deletions charts/gpu-base-operator-policy/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -5,19 +5,19 @@ Helm chart is for installing the Intel GPU base operator policy. The operator ha

## Helm install
```
helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-policy \
helm install --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-policy \
oci://ghcr.io/intel/intel-gpu-base-operator-policy-chart
Comment thread
tkatila marked this conversation as resolved.
```

## Helm upgrade
```
helm upgrade --namespace "intel-gpu-operator" --version 0.2.1 gpu-policy \
helm upgrade --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-policy \
oci://ghcr.io/intel/intel-gpu-base-operator-policy-chart
```

## Helm uninstall
```
helm uninstall --namespace "intel-gpu-operator" gpu-policy --wait
helm uninstall --namespace "intel-gpu-base-operator" gpu-policy --wait
```

## Configuration
Expand Down
4 changes: 2 additions & 2 deletions charts/gpu-base-operator/Chart.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -2,8 +2,8 @@ apiVersion: v2
name: intel-gpu-base-operator-chart
description: A Helm chart for Intel GPU Base Operator
type: application
version: 0.3.0-rc1
appVersion: "0.3.0-rc1"
version: 0.3.0
appVersion: "0.3.0"

dependencies:
- name: node-feature-discovery
Expand Down
11 changes: 3 additions & 8 deletions charts/gpu-base-operator/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -9,24 +9,19 @@ Helm chart is for installing the Intel GPU base operator. Operator installation

## Helm install
```
kubectl create ns intel-gpu-operator
# Required by DRA's admin access
kubectl label ns intel-gpu-operator resource.kubernetes.io/admin-access=true

helm install --namespace "intel-gpu-operator" --version 0.2.1 gpu-operator \
helm install --create-namespace --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-operator \
oci://ghcr.io/intel/intel-gpu-base-operator-chart --wait
```

## Helm upgrade
```
helm upgrade --namespace "intel-gpu-operator" --version 0.2.1 gpu-operator \
helm upgrade --namespace "intel-gpu-base-operator" --version 0.3.0 gpu-operator \
oci://ghcr.io/intel/intel-gpu-base-operator-chart --wait
```

## Helm uninstall
```
helm uninstall --namespace "intel-gpu-operator" gpu-operator
kubectl delete ns intel-gpu-operator
helm uninstall --namespace "intel-gpu-base-operator" gpu-operator
```

## Configuration
Expand Down
2 changes: 1 addition & 1 deletion charts/gpu-base-operator/values.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@ nfd:

operator:
image:
repository: ghcr.io/intel/intel-gpu-base-operator:0.3.0-rc1
repository: ghcr.io/intel/intel-gpu-base-operator:0.3.0
pullPolicy: IfNotPresent
verbosity: 2
resources:
Expand Down
2 changes: 1 addition & 1 deletion config/manager/manager.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -69,7 +69,7 @@ spec:
valueFrom:
fieldRef:
fieldPath: metadata.namespace
image: ghcr.io/intel/intel-gpu-base-operator:0.3.0-rc1
image: ghcr.io/intel/intel-gpu-base-operator:0.3.0
imagePullPolicy: IfNotPresent
name: manager
ports: []
Expand Down
2 changes: 1 addition & 1 deletion test/e2e/e2e_suite_test.go
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ var (

// projectImage is the name of the image which will be build and loaded
// with the code source changes to be tested.
projectImage = "ghcr.io/intel/intel-gpu-base-operator:0.3.0-rc1"
projectImage = "ghcr.io/intel/intel-gpu-base-operator:0.3.0"
)

// TestE2E runs the end-to-end (e2e) test suite for the project. These tests execute in an isolated,
Expand Down