Merge pull request #130 from canonical/main

Merge `main` into `track/0.10`
canonical · Jun 7, 2023 · fa5dbb0 · fa5dbb0
2 parents de140b0 + b1084ad
commit fa5dbb0
Show file tree

Hide file tree

Showing 36 changed files with 1,678 additions and 433 deletions.
diff --git a/.github/workflows/integrate.yaml b/.github/workflows/integrate.yaml
@@ -1,31 +1,28 @@
 # reusable workflow triggered by other actions
-name: Continuous Integration
+name: CI
 
 on:
   workflow_call:
     secrets:
-      charmcraft-credentials:
+      CHARMCRAFT_CREDENTIALS:
         required: true
 
 jobs:
-
   lib-check:
     name: Check libraries
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v3
-        with:
-          fetch-depth: 0
-      - name: Check libs
-        uses: canonical/charming-actions/[email protected]
-        with:
-          credentials: "${{ secrets.charmcraft-credentials }}"
-          github-token: "${{ secrets.GITHUB_TOKEN }}"
-
+    strategy:
+      matrix:
+        charm:
+        - kserve-controller
+        - kserve-web-app
+    uses: canonical/charmed-kubeflow-workflows/.github/workflows/_quality-checks.yaml@main
+    secrets: inherit
+    with:
+        charm-path: ./charms/${{ matrix.charm }}
+
   lint:
     name: Lint Code
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     strategy:
       matrix:
         # TODO: add web-app once changes in canonical/kserve-operators#2 and #3
@@ -38,7 +35,7 @@ jobs:
 
   unit:
     name: Unit Test
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     strategy:
       matrix:
         # TODO: add web-app once changes in canonical/kserve-operators#2 and #3
@@ -51,7 +48,7 @@ jobs:
 
   charm-integration:
     name: Individual Integration Tests
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     strategy:
       matrix:
         # TODO: add web-app once changes in canonical/kserve-operators#2 and #3
@@ -66,9 +63,10 @@ jobs:
         uses: charmed-kubernetes/actions-operator@main
         with:
           provider: microk8s
-          channel: 1.25/stable
+          channel: 1.24/stable
           # Pinned until this bug is resolved: https://bugs.launchpad.net/juju/+bug/1992833
           bootstrap-options: "--agent-version=2.9.34"
+          microk8s-addons: "dns storage rbac metallb:10.64.140.43-10.64.140.49"
           charmcraft-channel: latest/candidate
       - run: tox -e ${{ matrix.charm }}-integration
 
@@ -77,6 +75,14 @@ jobs:
         run: kubectl get all -A
         if: failure()
 
+      - name: Get inference servers
+        run: kubectl get inferenceservices -A
+        if: failure()
+
+      - name: Get events
+        run: kubectl get events -A
+        if: failure()
+
       - name: Describe deployments
         run: kubectl describe deployments -A
         if: failure()

diff --git a/.github/workflows/on_pull_request.yaml b/.github/workflows/on_pull_request.yaml
@@ -12,12 +12,11 @@ jobs:
   tests:
     name: Run Tests
     uses: ./.github/workflows/integrate.yaml
-    secrets:
-      charmcraft-credentials: ${{ secrets.CHARMCRAFT_CREDENTIALS }}
+    secrets: inherit
 
   # publish runs in parallel with tests, as we always publish in this situation
   publish-charm:
     name: Publish Charm
     uses: ./.github/workflows/publish.yaml
-    secrets:
-      CHARMCRAFT_CREDENTIALS: ${{ secrets.CHARMCRAFT_CREDENTIALS }}
+    secrets: inherit
+
diff --git a/.github/workflows/on_push.yaml b/.github/workflows/on_push.yaml
@@ -1,31 +1,28 @@
-name: Publish to edge if tests passed
+name: On Push
 
 # On push to a "special" branch, we:
 # * always publish to charmhub at latest/edge/branchname
 # * always run tests
-# where a "special" branch is one of main/master or track/**, as
+# where a "special" branch is one of main or track/**, as
 # by convention these branches are the source for a corresponding
 # charmhub edge channel.
 
 on:
   push:
     branches:
-      - master
-      - main
-      - track/**
+    - main
+    - track/**
 
 jobs:
 
   tests:
     name: Run Tests
     uses: ./.github/workflows/integrate.yaml
-    secrets:
-      charmcraft-credentials: "${{ secrets.CHARMCRAFT_CREDENTIALS }}"
+    secrets: inherit
 
   # publish runs in series with tests, and only publishes if tests passes
   publish-charm:
     name: Publish Charm
     needs: tests
     uses: ./.github/workflows/publish.yaml
-    secrets:
-      charmcraft-credentials: "${{ secrets.CHARMCRAFT_CREDENTIALS }}"
+    secrets: inherit
diff --git a/.github/workflows/publish.yaml b/.github/workflows/publish.yaml
@@ -58,7 +58,7 @@ jobs:
           ref: ${{ inputs.source_branch }}
 
       - name: Select charmhub channel
-        uses: canonical/charming-actions/channel@2.2.3
+        uses: canonical/charming-actions/channel@2.3.0
         id: select-channel
         if: ${{ inputs.destination_channel == '' }}
 
@@ -84,7 +84,7 @@ jobs:
           echo "::set-output name=tag_prefix::$tag_prefix"
 
       - name: Upload charm to charmhub
-        uses: canonical/charming-actions/upload-charm@2.2.3
+        uses: canonical/charming-actions/upload-charm@2.3.0
         with:
           credentials: ${{ secrets.CHARMCRAFT_CREDENTIALS }}
           github-token: ${{ secrets.GITHUB_TOKEN }}

diff --git a/.github/workflows/release.yaml b/.github/workflows/release.yaml
@@ -21,7 +21,7 @@ jobs:
     steps:
       - uses: actions/checkout@v2
       - name: Release charm to channel
-        uses: canonical/charming-actions/release-charm@2.2.3
+        uses: canonical/charming-actions/release-charm@2.3.0
         with:
           credentials: ${{ secrets.CHARMCRAFT_CREDENTIALS }}
           github-token: ${{ secrets.GITHUB_TOKEN }}

diff --git a/README.md b/README.md
@@ -7,12 +7,18 @@ Upstream documentation can be found at https://kserve.github.io/website/0.8/
 
 ## Usage
 
+### Supported versions
+
+* Kubernetes 1.24
+* istio-operators 1.16/stable
+* knative-operators 1.8/stable
+
 ### Pre-requisites
 
-* Microk8s <supported-version>/stable, supported-version=1.22, 1.23, 1.24
->NOTE: These instructions assume you have run `microk8s enable dns storage rbac metallb:"10.64.140.43-10.64.140.49,192.168.0.105-192.168.0.111"`
+* Kubernetes cluster
+>NOTE: If you are using Microk8s, it is assumed you have run `microk8s enable dns storage rbac metallb:"10.64.140.43-10.64.140.49,192.168.0.105-192.168.0.111"`.
 
-* `istio-pilot` and `istio-ingressgateway` charms deployed. See "Deploy dependencies" for deploy instructions.
+* `istio-pilot` and `istio-ingressgateway`. See "Deploy dependencies" for deploy instructions.
 
 ### Add a model and set variables
 
@@ -24,49 +30,87 @@ juju add-model ${MODEL_NAME}
 
 ### Deploy dependencies
 
-KServe requires istio to be deployed in the cluster. To correctly configure them, you can:
+kserve-operators require istio-operators to be deployed in the cluster. To correctly configure them, you can:
 
 ```
-ISTIO_CHANNEL=1.11/stable
+ISTIO_CHANNEL=1.16/stable
 juju deploy istio-pilot --config default-gateway=${DEFAULT_GATEWAY} --channel ${ISTIO_CHANNEL} --trust
 juju deploy istio-gateway istio-ingressgateway --config kind="ingress" --channel ${ISTIO_CHANNEL} --trust
 juju relate istio-pilot istio-ingressgateway
 ```
 
+#### Only for Serverless mode
+
+For serverless operations kserve-operators depends on knative-serving. To correctly configure it, you can:
+
+> NOTE: these instructions assume you have deployed Microk8s and MetalLB is enabled. If your cloud configuration is different than this, please refer to [knative-operators](https://github.com/canonical/knative-operators#usage) documentation.
+
+```
+KNATIVE_CHANNEL=1.8/stable
+juju deploy knative-operator --channel ${KNATIVE_CHANNEL} --trust
+juju deploy knative-serving --config namespace="knative-serving" --config istio.gateway.namespace=${MODEL_NAME} --config istio.gateway.name=${DEFAULT_GATEWAY} --channel ${KNATIVE_CHANNEL} --trust
+```
+
 ### Deploy in `RawDeployment` mode
 
-KServe supports `RawDeployment` mode to enable `InferenceService`, which removes the KNative dependency and unlocks some of its limitations, like mounting multiple volumes. Please note this mode is not loaded with serverless capabilities, for that you'd need to deploy in `Serverless` mode.
+kserve-operators support `RawDeployment` mode to manage `InferenceService`, which removes the KNative dependency and unlocks some of its limitations, like mounting multiple volumes. Please note this mode is not loaded with serverless capabilities, for that you'd need to deploy in `Serverless` mode.
 
 1. Deploy `kserver-controller`
 
 ```
 juju deploy kserve-controller --channel <channel> --trust
 ```
 
+2. Relate `kserve-controller` and `istio-pilot`
+
+```
+juju relate istio-pilot:gateway-info kserve-controller:ingress-gateway
+```
+
 > `channel` is the available channels of the Charmed KServe:
-* latest/stable
-* 0.8/stable
+* latest/edge
+* 0.10/stable
 
 ### Deploy in `Serverless` mode
 
-TODO
+kserve-operatos support `Serveless` mode to manage event driven `InferenceService`s, which enables autoscaling on demand, and supports scaling down to zero.
+
+1. Deploy `kserver-controller`
+
+```
+juju deploy kserve-controller --channel <channel> --config deployment-mode="serverless" --trust
+```
+
+2. Relate `kserve-controller` and `istio-pilot`
+
+```
+juju relate istio-pilot:gateway-info kserve-controller:ingress-gateway
+```
+
+3. Relate `kserve-controller` and `knative-serving`
+
+```
+juju relate kserve-controller:local-gateway knative-serving:local-gateway
+```
 
 ## Deploy a simple `InferenceService`
 
 To deploy a simple example of an `InferenceServer`, you can use the one provided in `examples/`
 
 > NOTE: this example is based on [First InferenceService](https://kserve.github.io/website/0.9/get_started/first_isvc/#2-create-an-inferenceservice)
 
-1. Create an `InferenceService`
+1. Create an `InferenceService` in a testing namespace
 
 ```
-kubectl apply -f sklearn-iris.yaml
+USER_NS="kserve-testing"
+kubectl create ns ${USER_NS}
+kubectl apply -f sklearn-iris.yaml -n${USER_NS}
 ```
 
 2. Check the `InferenceService` status
 
 ```
-kubectl get inferenceservices sklearn-iris
+kubectl get inferenceservices sklearn-iris -n${USER_NS}
 ```
 
 3. Determine the URL for performing inference
@@ -76,17 +120,16 @@ kubectl get inferenceservices sklearn-iris
 > NOTE: this method can only be used for performing inference within the cluster.
 
 ```
-SERVICE_IP=$(kubectl get svc sklearn-iris-predictor-default -ojsonpath='{.spec.clusterIP}')
+SERVICE_IP=$(kubectl get svc sklearn-iris-predictor-default -n${USER_NS} -ojsonpath='{.spec.clusterIP}')
 INFERENCE_URL="${SERVICE_IP}/v1/models/sklearn-iris:predict"
 ```
 
-* Using Istio Ingress and LoadBalancer
-
-> NOTE: This step assumes you enabled MetalLB addon in Microk8s at the beginning of this tutorial.
+* Using the `InferenceService` URL
 
 ```
-LOADBALANCER_IP=$(kubectl get svc istio-ingressgateway-workload -n${MODEL_NAME} -o jsonpath='{.status.loadBalancer.ingress[0].ip}')
-INFERENCE_URL="{LOADBALANCER_IP}/v1/models/sklearn-iris:predict"
+kubectl get inferenceservice sklearn-iris -n${USER_NS}
+# From the output, take the URL
+INFERENCE_URL="${URL}/v1/models/sklearn-iris:predict"
 ```
 
 4. Perform inference
@@ -102,6 +145,7 @@ cat <<EOF > "./iris-input.json"
   ]
 }
 EOF
+```
 
 Now call the `InferenceService`:
 

diff --git a/charms/kserve-controller/config.yaml b/charms/kserve-controller/config.yaml
@@ -1,6 +1,14 @@
 # Copyright 2022 Canonical Ltd.
 # See LICENSE file for licensing details.
 options:
+  deployment-mode:
+    default: "RawDeployment"
+    description: Deployment mode for kserve. It can only be RawDeployment or Serverless, and the latter requires knative-serving to be present in the model, and the local-gateway relation to be established.
+    type: string
+  domain-name:
+    default: "example.com"
+    description: The domain name used for all fully-qualified route names shown. In serverless mode, it has to be the same as knative-serving's domain name; otherwise, it can be changed to match the cluster ingress configuration.
+    type: string
   port:
     type: int
     default: 9443