Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly
Sign up
Appearance settings
This repository was archived by the owner on Jan 5, 2023. It is now read-only.

Archived: GitOps config repo for an Apollo GraphQL federated graph with a supergraph router and subgraph services deployed to Kubernetes.

License

Notifications You must be signed in to change notification settings

apollographql/supergraph-demo-k8s-graph-ops

Repository files navigation

Kubernetes-native GraphOps

Archived

The code in this repository is experimental and has been provided for reference purposes only. It is not actively maintained and has been archived.

Many folks have already migrated to Federation 2. If you're starting something new, please see the latest Federation docs!

Original Content

This is the GraphOps Repo for the apollographq/supergraph-demo Source Repo.

CI

GraphOps for Apollo Federation

Contents:

Welcome

Large-scale graph operators use Kubernetes to run their Graph Router and Subgraph Services, with continuous app and service delivery.

Kubernetes provides a mature control-plane for deploying and operating your graph using container images like those produced by the supergraph-demo Source Repo.

Overview

This repo follows the Declarative GitOps CD for Kubernetes Best Practices:

Source Repo - provides image tag versions via Bump image versions PR

  • apollographql/supergraph-demo produces the artifacts:
    • Subgraph docker images w/ embedded subgraph schemas
    • Supergraph-router docker image that can be fed a composed supergraph schema via
      • (a) Apollo Uplink - for update in place
      • (b) via a ConfigMap for declarative k8s config management
  • Continuous integration:
    • Bumps package version numbers & container tags.
    • Builds/publishes container images to container registry.
    • Bump image versions PRs to propagate image version bumps to this repo.

Graph Registry - provides supergraph schema via Bump supergraph schema PR

  • Graph schemas published to the Apollo Registry
    • Subgraphs publish their schema to the Apollo Registry after deployment.
    • Supergraph schema is published after Apollo Studio does:
      • Managed composition
      • Schema checks
      • Operation checks
      • No breaking changes are detected
    • Published supergraph schemas are made available via:
      • Apollo Uplink - that the Gateway can poll for live updates (default).
      • Apollo Registry - for retrieval via rover supergraph fetch.
      • Apollo Build Webhook - for triggering custom CD with the composed supergraph schema.
  • Bump supergraph schema PRs are created by the supergraph-build-webhook.yml workflow via:
    • Supergraph build webhook from Apollo Studio
    • Polling on a schedule in case a webhook is lost

GraphOps Repo (this repo) - declarative graph config for Kubernetes for GitOps

  • Declarative k8s configs for dev, stage, and prod
  • Promote config from dev -> stage -> prod
    • make promote-dev-stage
    • make promote-stage-prod
  • Continuous deployment:

kustomize for k8s-native config management:

Config Flow

Config data flows from the following sources:

  • Source Repo:

    • Bump image versions PR is opened on the GraphOps Repo:
      • when new Gateway docker image versions are published:
        • see end of the example artifact release workflow
        • bumps package versions in the Source Repo.
        • does an incremental monorepo build and pushes new docker images to DockerHub.
        • opens GraphOps Repo PR to bump the docker image versions in the dev environment (auto-merge).

  • Graph Registry:

    • Bump supergraph schema PR is opened on the GraphOps Repo:
      • when Managed Federation sends a supergraph schema build webhook
        • rover supergraph fetch is used to retrieve the supergraph schema from the Apollo Registry.

  • GraphOps Repo:

    • GraphOps team crafts the declarative configurations for each environment.
    • PRs for docker image bumps are (auto-)merged into the GraphOps Repo.
    • PR for supergraph schema bumps are (auto-)merged into the GraphOps Repo.
    • GraphOps Repo has definitive desired state for each environment.

Continous deployment of config data flows from the GraphOps Repo into the target k8s cluster:

  • Via kubectl apply -k or GitOps tools like ArgoCD or Flux

  • kustomize is used to generate parameterized config resources for each environment:

  • Progressive delivery controllers like Argo Rollouts or Flagger may also be used

    • BlueGreen and Canary deployment strategies

  • Rollback via git commit & GitOps or progressive delivery controller rollback.

Docker Images from Source Repo

New Gateway docker image versions are published as source changes are pushed to the main branch of the supergraph-demo repo.

This is done by the release.yml workflow, which does an incremental matrix build and pushes new docker images to DockerHub, and then opens a Bump image versions PR in this repo that uses kustomize edit set image to inject the new image version tags into the kustomization.yaml for each environment.

publish-artifacts-workflow

Note: This workflow can be easily adapted for a single-repo-per-package scenarios, where they separately publish their own docker images and issue separate version bump PRs to this GraphOps Repo.

Supergraph Schemas from Graph Registry

  1. Detecting changes to the supergraph built via Managed Federation

    • Managed Federation builds a supergraph schema after each rover subgraph publish
    • Changes detected with the following:
      • Supergraph build webhooks - when a new supergraph schema is built in Apollo Studio
      • rover supergraph fetch - to poll the Registry

  2. Bump supergraph schema PR with auto-merge enabled when changes detected

    • Workflow: supergraph-build-webhook.yml
    • Commits a new supergraph.graphql to the GraphOps Repo with the new version from Apollo Studio
    • Additional CI checks on the supergraph schema are required for the PR to merge
    • Auto-merged when CI checks pass

  3. Generate a new Gateway Deployment and ConfigMap using kustomize

    • Once changes to supergraph.graphql when Bump supergraph schema is merged

Using the Supergraph Build Webhook

  1. Register the webhook in Apollo Studio in your graph settings

    • Send the webhook to an automation service or serverless function:
    • webhook-register

  2. Adapt the webhook to a GitHub repository_dispatch POST request

  3. repository_dispatch event triggers a GitHub workflow

  4. GitHub workflow automatically creates a PR with auto-merge enabled

Deploy a Kubernetes Dev Environment

You'll need:

  • kubectl - with expanded kustomize support for resources
  • kind

then run:

make demo

which runs:

make k8s-up-dev

which creates:

  • local k8s cluster with the NGINX Ingress Controller
  • graph-router Deployment configured to use a supergraph ConfigMap
  • graph-router Service and Ingress

and applies the following:

kubectl apply -k infra/dev
kubectl apply -k subgraphs/dev
kubectl apply -k router/dev

Gateway Deployment with ConfigMap

using router/base/router.yaml:

apiVersion: apps/v1
kind: Deployment
metadata:
 labels:
 app: router
 name: router-deployment
spec:
 replicas: 1
 selector:
 matchLabels:
 app: router
 template:
 metadata:
 labels:
 app: router
 spec:
 containers:
 - env:
 - name: APOLLO_SCHEMA_CONFIG_EMBEDDED
 value: "true"
 image: prasek/supergraph-router:1.1.1
 name: router
 ports:
 - containerPort: 4000
 volumeMounts:
 - mountPath: /etc/config
 name: supergraph-volume
 volumes:
 - configMap:
 name: supergraph-c4mh62bddt
 name: supergraph-volume
---
apiVersion: v1
kind: ConfigMap
metadata:
 name: supergraph-c4mh62bddt
data:
 supergraph.graphql: |
 schema
 @core(feature: "https://specs.apollo.dev/core/v0.1"),
 @core(feature: "https://specs.apollo.dev/join/v0.1")
 {
 query: Query
 }

 directive @core(feature: String!) repeatable on SCHEMA

 directive @join__field(graph: join__Graph, requires: join__FieldSet, provides: join__FieldSet) on FIELD_DEFINITION

 directive @join__type(graph: join__Graph!, key: join__FieldSet) repeatable on OBJECT | INTERFACE

 directive @join__owner(graph: join__Graph!) on OBJECT | INTERFACE

 directive @join__graph(name: String!, url: String!) on ENUM_VALUE

 type DeliveryEstimates {
 estimatedDelivery: String
 fastestDelivery: String
 }

 scalar join__FieldSet

 enum join__Graph {
 INVENTORY @join__graph(name: "inventory" url: "http://inventory:4000/graphql")
 PRODUCTS @join__graph(name: "products" url: "http://products:4000/graphql")
 USERS @join__graph(name: "users" url: "https://users:4000/graphql")
 }

 type Product
 @join__owner(graph: PRODUCTS)
 @join__type(graph: PRODUCTS, key: "id")
 @join__type(graph: PRODUCTS, key: "sku package")
 @join__type(graph: PRODUCTS, key: "sku variation{id}")
 @join__type(graph: INVENTORY, key: "id")
 {
 id: ID! @join__field(graph: PRODUCTS)
 sku: String @join__field(graph: PRODUCTS)
 package: String @join__field(graph: PRODUCTS)
 variation: ProductVariation @join__field(graph: PRODUCTS)
 dimensions: ProductDimension @join__field(graph: PRODUCTS)
 createdBy: User @join__field(graph: PRODUCTS, provides: "totalProductsCreated")
 delivery(zip: String): DeliveryEstimates @join__field(graph: INVENTORY, requires: "dimensions{size weight}")
 }

 type ProductDimension {
 size: String
 weight: Float
 }

 type ProductVariation {
 id: ID!
 }

 type Query {
 allProducts: [Product] @join__field(graph: PRODUCTS)
 product(id: ID!): Product @join__field(graph: PRODUCTS)
 }

 type User
 @join__owner(graph: USERS)
 @join__type(graph: USERS, key: "email")
 @join__type(graph: PRODUCTS, key: "email")
 {
 email: ID! @join__field(graph: USERS)
 name: String @join__field(graph: USERS)
 totalProductsCreated: Int @join__field(graph: USERS)
 }
---
apiVersion: v1
kind: Service
metadata:
 name: router-service
spec:
 ports:
 - port: 4000
 protocol: TCP
 targetPort: 4000
 selector:
 app: router
---
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
 annotations:
 kubernetes.io/ingress.class: nginx
 name: router-ingress
spec:
 rules:
 - http:
 paths:
 - backend:
 service:
 name: router-service
 port:
 number: 4000
 path: /
 pathType: Prefix

and 3 subgraph services subgraphs/base/subgraphs.yaml:

make demo then runs the following in a loop until the query succeeds or 2 min timeout:

Make a GraphQL Query

kubectl get all
make k8s-query

which shows the following:

NAME READY STATUS RESTARTS AGE
pod/inventory-65494cbf8f-bhtft 1/1 Running 0 59s
pod/products-6d75ff449c-9sdnd 1/1 Running 0 59s
pod/router-deployment-84cbc9f689-8fcnf 1/1 Running 0 20s
pod/users-d85ccf5d9-cgn4k 1/1 Running 0 59s
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/inventory ClusterIP 10.96.108.120 <none> 4000/TCP 59s
service/kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 96s
service/products ClusterIP 10.96.65.206 <none> 4000/TCP 59s
service/router-service ClusterIP 10.96.178.206 <none> 4000/TCP 20s
service/users ClusterIP 10.96.98.53 <none> 4000/TCP 59s
NAME READY UP-TO-DATE AVAILABLE AGE
deployment.apps/inventory 1/1 1 1 59s
deployment.apps/products 1/1 1 1 59s
deployment.apps/router-deployment 1/1 1 1 20s
deployment.apps/users 1/1 1 1 59s
NAME DESIRED CURRENT READY AGE
replicaset.apps/inventory-65494cbf8f 1 1 1 59s
replicaset.apps/products-6d75ff449c 1 1 1 59s
replicaset.apps/router-deployment-84cbc9f689 1 1 1 20s
replicaset.apps/users-d85ccf5d9 1 1 1 59s
Smoke test
-------------------------------------------------------------------------------------------
++ curl -X POST -H 'Content-Type: application/json' --data '{ "query": "{ allProducts { id, sku, createdBy { email, totalProductsCreated } } }" }' http://localhost:80/
 % Total % Received % Xferd Average Speed Time Time Time Current
 Dload Upload Total Spent Left Speed
100 352 100 267 100 85 3000 955 --:--:-- --:--:-- --:--:-- 3911
{"data":{"allProducts":[{"id":"apollo-federation","sku":"federation","createdBy":{"email":"support@apollographql.com","totalProductsCreated":1337}},{"id":"apollo-studio","sku":"studio","createdBy":{"email":"support@apollographql.com","totalProductsCreated":1337}}]}}
Success!
-------------------------------------------------------------------------------------------

Cleanup

make demo then cleans up:

deployment.apps "graph-router" deleted
service "graphql-service" deleted
ingress.networking.k8s.io "graphql-ingress" deleted
Deleting cluster "kind" ...

Promoting to Stage and Prod

Promoting configs from from dev -> stage -> prod can be as simple as:

  1. copying the config from one environment to the next
    • make promote-dev-stage
    • make promote-stage-prod
  2. push the changes

The GitOps operator in each Kubernetes cluster will pull the environment configuration from this GraphOps Repo and any changes will be applied to that cluster.

GitOps

CD via GitOps:

We'll use flux v2 for this example, so you'll need:

then run:

make demo-flux

which runs:

make k8s-up-flux-dev

which shows something like:

.scripts/k8s-up-flux.sh dev
Using dev/kustomization.yaml
kind version 0.11.1
No kind clusters found.
+ kind create cluster --image kindest/node:v1.21.1 --config=clusters/kind-cluster.yaml --wait 5m
Creating cluster "kind" ...
 ✓ Ensuring node image (kindest/node:v1.21.1) 🖼
 ✓ Preparing nodes 📦
 ✓ Writing configuration 📜
 ✓ Starting control-plane 🕹️
 ✓ Installing CNI 🔌
 ✓ Installing StorageClass 💾
 ✓ Waiting ≤ 5m0s for control-plane = Ready ⏳
 • Ready after 28s 💚
Set kubectl context to "kind-kind"
You can now use your cluster with:
kubectl cluster-info --context kind-kind
Not sure what to do next? 😅 Check out https://kind.sigs.k8s.io/docs/user/quick-start/
+ flux install
✚ generating manifests
✔ manifests build completed
► installing components in flux-system namespace
◎にじゅうまる verifying installation
✔ source-controller: deployment ready
✔ kustomize-controller: deployment ready
✔ helm-controller: deployment ready
✔ notification-controller: deployment ready
✔ install finished
+ flux create source git k8s-graph-ops --url=https://github.com/apollographql/supergraph-demo-k8s-graph-ops.git --branch=main
✚ generating GitRepository source
► applying GitRepository source
✔ GitRepository source created
◎にじゅうまる waiting for GitRepository source reconciliation
✔ GitRepository source reconciliation completed
✔ fetched revision: main/13fbe62857a713f396947a552d0d72ca760d3010
+ flux create kustomization infra --namespace=default --source=GitRepository/k8s-graph-ops.flux-system --path=./infra/dev --prune=true --interval=1m --validation=client
✚ generating Kustomization
► applying Kustomization
✔ Kustomization created
◎にじゅうまる waiting for Kustomization reconciliation
✔ Kustomization infra is ready
✔ applied revision main/13fbe62857a713f396947a552d0d72ca760d3010
+ flux create kustomization subgraphs --namespace=default --source=GitRepository/k8s-graph-ops.flux-system --path=./subgraphs/dev --prune=true --interval=1m --validation=client
✚ generating Kustomization
► applying Kustomization
✔ Kustomization created
◎にじゅうまる waiting for Kustomization reconciliation
✔ Kustomization subgraphs is ready
✔ applied revision main/13fbe62857a713f396947a552d0d72ca760d3010
+ flux create kustomization router --depends-on=infra --namespace=default --source=GitRepository/k8s-graph-ops.flux-system --path=./router/dev --prune=true --interval=1m --validation=client
✚ generating Kustomization
► applying Kustomization
✔ Kustomization created
◎にじゅうまる waiting for Kustomization reconciliation

the router ingress config needs the nginx ingress controller, so you'll see this while the nginx ingress admission controller is starting, but with GitOps and flux the configuration will be re-applied so will self-heal once it's started:

✗ apply failed: Error from server (InternalError): error when creating "3a946b48-8ea1-4516-8dcf-7341332f4d88.yaml": Internal error occurred: failed calling webhook "validate.nginx.ingress.kubernetes.io": Post "https://ingress-nginx-controller-admission.ingress-nginx.svc:443/networking/v1beta1/ingresses?timeout=10s": dial tcp 10.96.26.139:443: i/o timeout

then smoke tests will run while the nginx ingress admission controller is starting, with the initial tests failing while the nginx admission controller is starting:

-------------------------
Test 1
-------------------------
++ curl -X POST -H 'Content-Type: application/json' --data '{ "query": "{allProducts{delivery{estimatedDelivery,fastestDelivery},createdBy{name,email}}}" }' http://localhost:80/
 % Total % Received % Xferd Average Speed Time Time Time Current
 Dload Upload Total Spent Left Speed
100 241 100 146 100 95 29200 19000 --:--:-- --:--:-- --:--:-- 48200
-------------------------
❌ Test 1
-------------------------
[Expected]
{"data":{"allProducts":[{"delivery":{"estimatedDelivery":"6/25/2021","fastestDelivery":"6/24/2021"},"createdBy":{"name":"Apollo Studio Support","email":"support@apollographql.com"}},{"delivery":{"estimatedDelivery":"6/25/2021","fastestDelivery":"6/24/2021"},"createdBy":{"name":"Apollo Studio Support","email":"support@apollographql.com"}}]}}
-------------------------
[Actual]
<html>
<head><title>404 Not Found</title></head>
<body>
<center><h1>404 Not Found</h1></center>
<hr><center>nginx</center>
</body>
</html>
-------------------------
❌ Test 1
-------------------------

and then once it's started and the router ingress can be applied, the smoke tests will pass:

-------------------------
Test 1
-------------------------
++ curl -X POST -H 'Content-Type: application/json' --data '{ "query": "{allProducts{delivery{estimatedDelivery,fastestDelivery},createdBy{name,email}}}" }' http://localhost:80/
 % Total % Received % Xferd Average Speed Time Time Time Current
 Dload Upload Total Spent Left Speed
100 438 100 343 100 95 1982 549 --:--:-- --:--:-- --:--:-- 2531
Result:
{"data":{"allProducts":[{"delivery":{"estimatedDelivery":"6/25/2021","fastestDelivery":"6/24/2021"},"createdBy":{"name":"Apollo Studio Support","email":"support@apollographql.com"}},{"delivery":{"estimatedDelivery":"6/25/2021","fastestDelivery":"6/24/2021"},"createdBy":{"name":"Apollo Studio Support","email":"support@apollographql.com"}}]}}
✅ Test 1
-------------------------
Test 2
-------------------------
++ curl -X POST -H 'Content-Type: application/json' --data '{ "query": "{allProducts{id,sku,createdBy{email,totalProductsCreated}}}" }' http://localhost:80/
 % Total % Received % Xferd Average Speed Time Time Time Current
 Dload Upload Total Spent Left Speed
100 341 100 267 100 74 17800 4933 --:--:-- --:--:-- --:--:-- 22733
Result:
{"data":{"allProducts":[{"id":"apollo-federation","sku":"federation","createdBy":{"email":"support@apollographql.com","totalProductsCreated":1337}},{"id":"apollo-studio","sku":"studio","createdBy":{"email":"support@apollographql.com","totalProductsCreated":1337}}]}}
✅ Test 2
✅ All tests pass! 🚀

then finally the kind cluster will be deleted:

.scripts/k8s-down.sh
Deleting cluster "kind" ...

Progressive Delivery

See the BlueGreen example below with more advanced examples coming soon!

BlueGreen Deploys with Argo Rollouts

We'll use Argo Rollouts to do a basic BlueGreen deployment in this example.

Initial Deployment

make k8s-up-flux-bluegreen

which does does a BlueGreen deploy of the subgraphs using GitOps and subgraphs/dev-bluegreen/kustomization.yaml

and shows the following:

.scripts/k8s-up-flux.sh dev bluegreen
Using Kustomizations:
- infra/dev/kustomization.yaml
- subgraphs/dev-bluegreen/kustomization.yaml
- router/dev/kustomization.yaml
kind version 0.11.1
No kind clusters found.
Creating cluster "kind" ...
 ✓ Ensuring node image (kindest/node:v1.21.1) 🖼
 ✓ Preparing nodes 📦
 ✓ Writing configuration 📜
 ✓ Starting control-plane 🕹️
 ✓ Installing CNI 🔌
 ✓ Installing StorageClass 💾
 ✓ Waiting ≤ 5m0s for control-plane = Ready ⏳
 • Ready after 28s 💚
Set kubectl context to "kind-kind"
You can now use your cluster with:
kubectl cluster-info --context kind-kind
Thanks for using kind! 😊
+ flux install
✚ generating manifests
✔ manifests build completed
► installing components in flux-system namespace
◎にじゅうまる verifying installation
✔ notification-controller: deployment ready
✔ source-controller: deployment ready
✔ kustomize-controller: deployment ready
✔ helm-controller: deployment ready
✔ install finished
+ flux create source git k8s-graph-ops --url=https://github.com/apollographql/supergraph-demo-k8s-graph-ops.git --branch=main
✚ generating GitRepository source
► applying GitRepository source
✔ GitRepository source created
◎にじゅうまる waiting for GitRepository source reconciliation
✔ GitRepository source reconciliation completed
✔ fetched revision: main/9c6b88c18faecc76047a75e842f837c00d79f1f4
+ flux create kustomization infra --namespace=default --source=GitRepository/k8s-graph-ops.flux-system --path=./infra/dev --prune=true --interval=1m --validation=client
✚ generating Kustomization
► applying Kustomization
✔ Kustomization created
◎にじゅうまる waiting for Kustomization reconciliation
✔ Kustomization infra is ready
✔ applied revision main/9c6b88c18faecc76047a75e842f837c00d79f1f4
+ flux create kustomization subgraphs --namespace=default --source=GitRepository/k8s-graph-ops.flux-system --path=./subgraphs/dev-bluegreen --prune=true --interval=1m --validation=client
✚ generating Kustomization
► applying Kustomization
✔ Kustomization created
◎にじゅうまる waiting for Kustomization reconciliation
✔ Kustomization subgraphs is ready
✔ applied revision main/9c6b88c18faecc76047a75e842f837c00d79f1f4
+ flux create kustomization router --depends-on=infra --namespace=default --source=GitRepository/k8s-graph-ops.flux-system --path=./router/dev --prune=true --interval=1m --validation=client
✚ generating Kustomization
► applying Kustomization
✔ Kustomization created
◎にじゅうまる waiting for Kustomization reconciliation
✗ apply failed: Error from server (InternalError): error when creating "76b5f11b-1666-48d5-80ca-862d183f2248.yaml": Internal error occurred: failed calling webhook "validate.nginx.ingress.kubernetes.io": Post "https://ingress-nginx-controller-admission.ingress-nginx.svc:443/networking/v1beta1/ingresses?timeout=10s": context deadline exceeded
+ kubectl wait --namespace ingress-nginx --for=condition=ready pod --selector=app.kubernetes.io/component=controller --timeout=120s
pod/ingress-nginx-controller-6cd89dbf45-sjs49 condition met

Verify Initial Deployment

you can then run:

make smoke

which shows the following:

.scripts/k8s-smoke.sh
-------------------------
Test 1
-------------------------
++ curl -X POST -H 'Content-Type: application/json' --data '{ "query": "{allProducts{delivery{estimatedDelivery,fastestDelivery},createdBy{name,email}}}" }' http://localhost:80/
 % Total % Received % Xferd Average Speed Time Time Time Current
 Dload Upload Total Spent Left Speed
100 438 100 343 100 95 1366 378 --:--:-- --:--:-- --:--:-- 1745
Result:
{"data":{"allProducts":[{"delivery":{"estimatedDelivery":"6/25/2021","fastestDelivery":"6/24/2021"},"createdBy":{"name":"Apollo Studio Support","email":"support@apollographql.com"}},{"delivery":{"estimatedDelivery":"6/25/2021","fastestDelivery":"6/24/2021"},"createdBy":{"name":"Apollo Studio Support","email":"support@apollographql.com"}}]}}
✅ Test 1
-------------------------
Test 2
-------------------------
++ curl -X POST -H 'Content-Type: application/json' --data '{ "query": "{allProducts{id,sku,createdBy{email,totalProductsCreated}}}" }' http://localhost:80/
 % Total % Received % Xferd Average Speed Time Time Time Current
 Dload Upload Total Spent Left Speed
100 341 100 267 100 74 14052 3894 --:--:-- --:--:-- --:--:-- 17947
Result:
{"data":{"allProducts":[{"id":"apollo-federation","sku":"federation","createdBy":{"email":"support@apollographql.com","totalProductsCreated":1337}},{"id":"apollo-studio","sku":"studio","createdBy":{"email":"support@apollographql.com","totalProductsCreated":1337}}]}}
✅ Test 2
✅ All tests pass! 🚀

using these Rollouts:

kubectl get rollouts
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE
inventory-bluegreen 1 1 1 1
products-bluegreen 1 1 1 1
users-bluegreen 1 1 1 1

and these Kustomizations:

kubectl get kustomization
NAME READY STATUS AGE
infra True Applied revision: main/9c6b88c18faecc76047a75e842f837c00d79f1f4 2m9s
router True Applied revision: main/9c6b88c18faecc76047a75e842f837c00d79f1f4 107s
subgraphs True Applied revision: main/9c6b88c18faecc76047a75e842f837c00d79f1f4 109s

Make a Change to the Products Subgraph

Pushing a subgraph change to the products subgraph in the supergraph-demo results in:

kubectl get kustomization
NAME READY STATUS AGE
infra True Applied revision: main/b4e5b385ac6ddb145cf2f95f77bda678997c75e4 78m
router True Applied revision: main/b4e5b385ac6ddb145cf2f95f77bda678997c75e4 78m
subgraphs True Applied revision: main/b4e5b385ac6ddb145cf2f95f77bda678997c75e4 78m

Preview Deployment Created

  • the Argo Rollouts controller then deploys a new bluegreen deployment and makes it available via the product preview-service.

Which shows the following:

kubectl get rollouts
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE
inventory-bluegreen 1 1 1 1
products-bluegreen 1 2 1 1
users-bluegreen 1 1 1 1
kubectl get pods
NAME READY STATUS RESTARTS AGE
pod/inventory-bluegreen-59d479fc9f-57fqw 1/1 Running 0 78m
pod/products-bluegreen-599c9f6c88-k7dwx 1/1 Running 0 78m
pod/products-bluegreen-6fb56d84ff-k2jks 1/1 Running 0 67m
pod/router-deployment-588b77bc9b-k9gz5 1/1 Running 0 78m
pod/users-bluegreen-6b789d8cb7-wrxt7 1/1 Running 0 78m

Promote Preview Deployment to Active

Since we've set the product subgraph rollout to have:

 # Rollouts can be resumed using: `kubectl argo rollouts promote ROLLOUT`
 autoPromotionEnabled: false

we can install and use the Argo Rollouts Kubectl Plugin to manually promote the BlueGreen deployment for the product service:

kubectl argo rollouts promote products-bluegreen
rollout 'products-bluegreen' promoted

which results in the preview products deployment becoming active and the previous active deployment being decomissioned, resulting in one active pod and replicaset for the products subgraph.

kubectl get rollouts
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE
inventory-bluegreen 1 1 1 1
products-bluegreen 1 1 1 1
users-bluegreen 1 1 1 1

and

kubectl get pods
NAME READY STATUS RESTARTS AGE
pod/inventory-bluegreen-59d479fc9f-57fqw 1/1 Running 0 80m
pod/products-bluegreen-6fb56d84ff-k2jks 1/1 Running 0 69m
pod/router-deployment-588b77bc9b-k9gz5 1/1 Running 0 80m
pod/users-bluegreen-6b789d8cb7-wrxt7 1/1 Running 0 80m

Verify new Active Deployment

make smoke

which shows:

-------------------------
Test 1
-------------------------
++ curl -X POST -H 'Content-Type: application/json' --data '{ "query": "{allProducts{delivery{estimatedDelivery,fastestDelivery},createdBy{name,email}}}" }' http://localhost:80/
 % Total % Received % Xferd Average Speed Time Time Time Current
 Dload Upload Total Spent Left Speed
100 438 100 343 100 95 641 177 --:--:-- --:--:-- --:--:-- 818
Result:
{"data":{"allProducts":[{"delivery":{"estimatedDelivery":"6/25/2021","fastestDelivery":"6/24/2021"},"createdBy":{"name":"Apollo Studio Support","email":"support@apollographql.com"}},{"delivery":{"estimatedDelivery":"6/25/2021","fastestDelivery":"6/24/2021"},"createdBy":{"name":"Apollo Studio Support","email":"support@apollographql.com"}}]}}
✅ Test 1
-------------------------
Test 2
-------------------------
++ curl -X POST -H 'Content-Type: application/json' --data '{ "query": "{allProducts{id,sku,createdBy{email,totalProductsCreated}}}" }' http://localhost:80/
 % Total % Received % Xferd Average Speed Time Time Time Current
 Dload Upload Total Spent Left Speed
100 341 100 267 100 74 22250 6166 --:--:-- --:--:-- --:--:-- 28416
Result:
{"data":{"allProducts":[{"id":"apollo-federation","sku":"federation","createdBy":{"email":"support@apollographql.com","totalProductsCreated":1337}},{"id":"apollo-studio","sku":"studio","createdBy":{"email":"support@apollographql.com","totalProductsCreated":1337}}]}}
✅ Test 2
✅ All tests pass! 🚀

Cleanup

make k8s-down

which shows

.scripts/k8s-down.sh
Deleting cluster "kind" ...

Learn More

Checkout the apollographq/supergraph-demo Source Repo.

Learn more about how Apollo can help your teams ship faster.

About

Archived: GitOps config repo for an Apollo GraphQL federated graph with a supergraph router and subgraph services deployed to Kubernetes.

Topics

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Contributors 2

AltStyle によって変換されたページ (->オリジナル) /