A simple trading application that runs in the public cloud but keeps its data in a private Kafka cluster
This example is part of a suite of examples showing the different ways you can use Skupper to connect services across cloud providers, data centers, and edge sites.
- Overview
- Prerequisites
- Step 1: Install the Skupper command-line tool
- Step 2: Set up your namespaces
- Step 3: Deploy the Kafka cluster
- Step 4: Deploy the application services
- Step 5: Create your sites
- Step 6: Link your sites
- Step 7: Expose the Kafka cluster
- Step 8: Access the frontend
- Cleaning up
- Summary
- Next steps
- About this example
This example is a simple Kafka application that shows how you can use Skupper to access a Kafka cluster at a remote site without exposing it to the public internet.
It contains four services:
-
A Kafka cluster running in a private data center. The cluster has two topics, "orders" and "updates".
-
An order processor running in the public cloud. It consumes from "orders", matching buy and sell offers to make trades. It publishes new and updated orders and trades to "updates".
-
A market data service running in the public cloud. It looks at the completed trades and computes the latest and average prices, which it then publishes to "updates".
-
A web frontend service running in the public cloud. It submits buy and sell orders to "orders" and consumes from "updates" in order to show what's happening.
To set up the Kafka cluster, this example uses the Kubernetes operator from the Strimzi project. The other services are small Python programs.
The example uses two Kubernetes namespaces, "private" and "public", to represent the private data center and public cloud.
-
The
kubectl
command-line tool, version 1.15 or later (installation guide) -
Access to at least one Kubernetes cluster, from any provider you choose
This example uses the Skupper command-line tool to deploy Skupper.
You need to install the skupper
command only once for each
development environment.
On Linux or Mac, you can use the install script (inspect it here) to download and extract the command:
curl https://skupper.io/install.sh | sh
The script installs the command under your home directory. It prompts you to add the command to your path if necessary.
For Windows and other installation options, see Installing Skupper.
Skupper is designed for use with multiple Kubernetes namespaces,
usually on different clusters. The skupper
and kubectl
commands use your kubeconfig and current context to
select the namespace where they operate.
Your kubeconfig is stored in a file in your home directory. The
skupper
and kubectl
commands use the KUBECONFIG
environment
variable to locate it.
A single kubeconfig supports only one active context per user. Since you will be using multiple contexts at once in this exercise, you need to create distinct kubeconfigs.
For each namespace, open a new terminal window. In each terminal,
set the KUBECONFIG
environment variable to a different path and
log in to your cluster. Then create the namespace you wish to use
and set the namespace on your current context.
Note: The login procedure varies by provider. See the documentation for yours:
- Minikube
- Amazon Elastic Kubernetes Service (EKS)
- Azure Kubernetes Service (AKS)
- Google Kubernetes Engine (GKE)
- IBM Kubernetes Service
- OpenShift
Public:
export KUBECONFIG=~/.kube/config-public
# Enter your provider-specific login command
kubectl create namespace public
kubectl config set-context --current --namespace public
Private:
export KUBECONFIG=~/.kube/config-private
# Enter your provider-specific login command
kubectl create namespace private
kubectl config set-context --current --namespace private
In Private, use the kubectl create
and kubectl apply
commands with the listed YAML files to install the operator and
deploy the cluster and topic.
Private:
kubectl create -f kafka-cluster/strimzi.yaml
kubectl apply -f kafka-cluster/cluster1.yaml
kubectl wait --for condition=ready --timeout 900s kafka/cluster1
Note:
By default, the Kafka bootstrap server returns broker addresses that include the Kubernetes namespace in their domain name. When, as in this example, the Kafka client is running in a namespace with a different name from that of the Kafka cluster, this prevents the client from resolving the Kafka brokers.
To make the Kafka brokers reachable, set the advertisedHost
property of each broker to a domain name that the Kafka client
can resolve at the remote site. In this example, this is
achieved with the following listener configuration:
spec:
kafka:
listeners:
- name: plain
port: 9092
type: internal
tls: false
configuration:
brokers:
- broker: 0
advertisedHost: cluster1-kafka-0.cluster1-kafka-brokers
See Advertised addresses for brokers for more information.
In Public, use the kubectl apply
command with the listed YAML
files to install the application services.
Public:
kubectl apply -f order-processor/kubernetes.yaml
kubectl apply -f market-data/kubernetes.yaml
kubectl apply -f frontend/kubernetes.yaml
A Skupper site is a location where components of your application are running. Sites are linked together to form a network for your application. In Kubernetes, a site is associated with a namespace.
For each namespace, use skupper init
to create a site. This
deploys the Skupper router and controller. Then use skupper status
to see the outcome.
Note: If you are using Minikube, you need to start minikube
tunnel before you run skupper init
.
Public:
skupper init
skupper status
Sample output:
$ skupper init
Waiting for LoadBalancer IP or hostname...
Waiting for status...
Skupper is now installed in namespace 'public'. Use 'skupper status' to get more information.
$ skupper status
Skupper is enabled for namespace "public". It is not connected to any other sites. It has no exposed services.
Private:
skupper init
skupper status
Sample output:
$ skupper init
Waiting for LoadBalancer IP or hostname...
Waiting for status...
Skupper is now installed in namespace 'private'. Use 'skupper status' to get more information.
$ skupper status
Skupper is enabled for namespace "private". It is not connected to any other sites. It has no exposed services.
As you move through the steps below, you can use skupper status
at
any time to check your progress.
A Skupper link is a channel for communication between two sites. Links serve as a transport for application connections and requests.
Creating a link requires use of two skupper
commands in
conjunction, skupper token create
and skupper link create
.
The skupper token create
command generates a secret token that
signifies permission to create a link. The token also carries the
link details. Then, in a remote site, The skupper link create
command uses the token to create a link to the site
that generated it.
Note: The link token is truly a secret. Anyone who has the token can link to your site. Make sure that only those you trust have access to it.
First, use skupper token create
in site Public to generate the
token. Then, use skupper link create
in site Private to link
the sites.
Public:
skupper token create ~/secret.token
Sample output:
$ skupper token create ~/secret.token
Token written to ~/secret.token
Private:
skupper link create ~/secret.token
Sample output:
$ skupper link create ~/secret.token
Site configured to link to https://10.105.193.154:8081/ed9c37f6-d78a-11ec-a8c7-04421a4c5042 (name=link1)
Check the status of the link using 'skupper link status'.
If your terminal sessions are on different machines, you may need
to use scp
or a similar tool to transfer the token securely. By
default, tokens expire after a single use or 15 minutes after
creation.
In Private, use skupper expose
with the --headless
option to
expose the Kafka cluster as a headless service on the Skupper
network.
Then, in Public, use kubectl get service
to check that the
cluster1-kafka-brokers
service appears after a moment.
Private:
skupper expose statefulset/cluster1-kafka --headless --port 9092
Public:
kubectl get service/cluster1-kafka-brokers
In order to use and test the application, we need external access to the frontend.
Use kubectl expose
with --type LoadBalancer
to open network
access to the frontend service.
Once the frontend is exposed, use kubectl get service/frontend
to look up the external IP of the frontend service. If the
external IP is <pending>
, try again after a moment.
Once you have the external IP, use curl
or a similar tool to
request the /api/health
endpoint at that address.
Note: The <external-ip>
field in the following commands is a
placeholder. The actual value is an IP address.
Public:
kubectl expose deployment/frontend --port 8080 --type LoadBalancer
kubectl get service/frontend
curl http://<external-ip>:8080/api/health
Sample output:
$ kubectl expose deployment/frontend --port 8080 --type LoadBalancer
service/frontend exposed
$ kubectl get service/frontend
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
frontend LoadBalancer 10.103.232.28 <external-ip> 8080:30407/TCP 15s
$ curl http://<external-ip>:8080/api/health
OK
If everything is in order, you can now access the web interface by
navigating to http://<external-ip>:8080/
in your browser.
To remove Skupper and the other resources from this exercise, use the following commands.
Private:
skupper delete
kubectl delete -f kafka-cluster/cluster1.yaml
kubectl delete -f kafka-cluster/strimzi.yaml
Public:
skupper delete
kubectl delete -f frontend/kubernetes.yaml
kubectl delete -f market-data/kubernetes.yaml
kubectl delete -f order-processor/kubernetes.yaml
Check out the other examples on the Skupper website.
This example was produced using Skewer, a library for documenting and testing Skupper examples.
Skewer provides utility functions for generating the README and
running the example steps. Use the ./plano
command in the project
root to see what is available.
To quickly stand up the example using Minikube, try the ./plano demo
command.