Add basic release workflow

.github/workflows/release.yaml

@@ -0,0 +1,90 @@
+name: release
+on:
+  push:
+    tags:
+      - "{,v}*.*.*"
+
+env:
+  AGENT_IMAGE: "neondatabase/autoscaler-agent"
+  SCHED_IMAGE: "neondatabase/autoscale-scheduler"
+  # pending merge of #8:
+  # INFORMANT_IMAGE: "neondatabase/vm-informant"
+
+jobs:
+  release:
+    runs-on: ubuntu-latest
+    steps:
+      - name: git checkout
+        uses: actions/checkout@v3
+      - name: get version and git info
+        id: get_vcs_info
+        run: |
+          echo "version=${{ github.ref_name }}" >> $GITHUB_OUTPUT
+          echo -n "git_info=" >> $GITHUB_OUTPUT
+          bash -c "set -eu -o pipefail; source ./scripts-common.sh; git_info" >> $GITHUB_OUTPUT
+      - name: install golang
+        uses: actions/setup-go@v3
+        with:
+          go-version-file: 'go.mod'
+          cache: true
+
+      - name: docker - install qemu
+        uses: docker/setup-qemu-action@v2
+      - name: docker - setup buildx
+        uses: docker/setup-buildx-action@v2
+      - name: login to docker hub
+        uses: docker/login-action@v2
+        with:
+          username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
+          password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
+
+      - name: check everything builds
+        run: go build ./...
+
+      - name: build and push autoscale-scheduler image
+        uses: docker/build-push-action@v3
+        with:
+          context: .
+          push: true
+          file: build/autoscale-scheduler/Dockerfile
+          tags: ${{ env.SCHED_IMAGE }}:${{ steps.get_vcs_info.outputs.version }}
+          build-args:
+            - GIT_INFO=${{ setps.get_vcs_info.outputs.git_info }}
+      - name: build and push autoscaler-agent image
+        uses: docker/build-push-action@v3
+        with:
+          context: .
+          push: true
+          file: build/autoscaler-agent/Dockerfile
+          tags: ${{ env.SCHED_IMAGE }}:${{ steps.get_vcs_info.outputs.version }}
+          build-args:
+            - GIT_INFO=${{ setps.get_vcs_info.outputs.git_info }}
+      # pending merge of #8:
+      # - name: build and push vm-informant image
+      #   uses: docker/build-push-action@v3
+      #   with:
+      #     context: .
+      #     push: true
+      #     file: build/vm-informant/Dockerfile
+      #     tags: ${{ env.INFORMANT_IMAGE }}:${{ steps.get_vcs_info.outputs.version }}
+      #     build-args:
+      #       - GIT_INFO=${{ setps.get_vcs_info.outputs.git_info }}
+      #
+      # # Because we want a docker image for the VM informant, the easiest way for us to also provide
+      # # a binary is by just extracting it from the container image itself.
+      # - name: extract vm-informant binary
+      #   run: |
+      #     mkdir bin
+      #     # todo: do we actually need to pull here if we're pushing above?
+      #     docker pull --quiet ${{ env.INFORMANT_IMAGE }}:${{ steps.get_vcs_info.outputs.version }}
+      #     ID=$(docker create ${{ env.INFORMANT_IMAGE }}:${{ steps.get_vcs_info.outputs.version }}
+      #     docker cp $ID:/usr/bin/vm-informant bin/vm-informant
+      #     docker rm -f $ID
+
+      - name: github release
+        uses: softprops/action-gh-release@v1
+        with:
+          files: |
+            bin/vm-informant
+            deploy/autoscaler-agent.yaml
+            deploy/autsocale-scheduler.yaml

ARCHITECTURE.md

@@ -70,8 +70,8 @@ discussed more in the [high-level consequences] section below.
     functionality implemented here. (See: `pkg/agent`, `pkg/informant`, and `pkg/plugin`)
 * `deploy/` — YAML files used during cluster init. Of these, only the following two are manually
   written:
-    * `deploy/autoscaler-agent-deploy.yaml`
-    * `deploy/scheduler-deploy.yaml`
+    * `deploy/autoscaler-agent.yaml`
+    * `deploy/autoscale-scheduler.yaml`
 * `kind/` — files specific to creating our [kind](https://kind.sigs.k8s.io/) cluster
     * `kind/cni-bin/` — [CNI plugins](https://www.cni.dev/docs/) downloaded via
       [`scripts/download-cni.sh`](scripts/download-cni.sh), required for our bridge network.

README.md

@@ -2,6 +2,18 @@
 
 Vertical autoscaling for a fleet of postgres instances running in a Kubernetes cluster.
 
+## Quick access
+
+Images are available as:
+
+| Component name | Image name |
+|----------------|------------|
+| scheduler (and plugin) | `neondatabase/autoscale-scheduler` |
+| autoscaler-agent | `neondatabase/autoscaler-agent` |
+| VM informant | `neondatabase/vm-informant` |
+
+The deployment files and a VM informant binary are attached to each release.
+
 ## Overview
 
 We want to dynamically change the amount of CPUs and memory of running postgres instances, _without

scripts/cluster-init.sh

@@ -27,7 +27,7 @@ kubectl apply -f deploy/neonvm.yaml | indent
 
 kubectl wait deployment -n neonvm-system neonvm-controller --for=condition=Available=True | indent
 
-kubectl apply -f deploy/scheduler-deploy.yaml -f deploy/autoscaler-agent-deploy.yaml | indent
+kubectl apply -f deploy/autoscale-scheduler.yaml -f deploy/autoscaler-agent.yaml | indent
 
 kubectl create secret generic vm-ssh --from-file=private-key=vm_image/ssh_id_rsa | indent
 

scripts/replace-scheduler.sh

@@ -15,4 +15,4 @@ if [ -n "$pod" ]; then
     kubectl delete pod -n kube-system "$pod"
 fi
 
-kubectl apply -f deploy/scheduler-deploy.yaml
+kubectl apply -f deploy/autoscale-scheduler.yaml