This post was originally published on Yugabyte’s blog.
Spring Boot is one of the most popular frameworks for building cloud native applications. It makes configuring an application easy and offers tons of starters to get you off the ground quickly. Each Spring Boot application is stand-alone and self-contained, which makes them easy to deploy in a distributed fashion – to containers or, even better, on Kubernetes.
Yugabyte Cloud is a perfect match for Spring Boot applications, especially ones made highly available with Kubernetes. Yugabyte Cloud gives you a PostgreSQL-compatible database that is resilient and scalable, just like the applications you run on k8s. You can scale your YugabyteDB cluster out (or in) with just a few clicks. So you can spend less time worrying about running your database and more time focused on building better software. It’s cloud native SQL for cloud native applications. Sign up for Yugabyte Cloud today.
In this post, we’ll walk you through connecting a Spring Boot application to Yugabyte Cloud and deploying it on minikube. You can view our walkthrough on connecting a Spring Boot application to Yugabyte Cloud and deploying it on Kubernetes via Amazon Elastic Kubernetes Service (EKS) here and via Google Kubernetes Engine (GKE) here. We’ll be using a slightly updated version of the popular Spring Boot PetClinic sample application that has a profile making it compatible with YugabyteDB. The repository for this is at https://github.com/yugabyte/spring-petclinic.
In this walkthrough, you will:
- Create a free Yugabyte Cloud account and database cluster
- Download the Spring Boot PetClinic sample application and connect it to Yugabyte Cloud
- Containerize the Spring Boot PetClinic sample application
- Start minikube
- Deploy the Spring Boot PetClinic sample application image to minikube
Note: Anything in brackets [ ] needs to be replaced with information from your deployment.
Create a free Yugabyte Cloud account and database cluster
Click the Create a Free Cluster button.
Select Yugabyte Cloud Free and click the Next button.
Select a cloud provider for your cluster, name your cluster, select a region, and click the Next button.
Copy the default admin credentials by clicking the Copy Credentials link. Then check the “I have copied the admin credentials” checkbox and click the Create Cluster button.
a. You can also set your own admin credentials by selecting the “Add your own credentials” radio button and entering your own credentials.
Wait for your cluster to be provisioned. Click on “Add IP Allow List” in the top section of the Clusters page.
Click the Create New List and Add to Cluster button.
In the Name text box, enter “all-ips”. In the IP Address(es) or Range text box, enter “0.0.0.0/0”. Click the Save button. This will allow traffic from all IP addresses to your cluster.
a. Allowing all IPs access to your cluster is for development purposes only. When you deploy an application to Production, you will want to specify only the specific IP addresses or ranges that need access to your cluster.
Your cluster is now created in Yugabyte Cloud and is accessible to all IP addresses. Leave this page open, as you’ll be accessing the Cloud Shell later to create an application database and user.
Download the Spring Boot PetClinic sample application and connect it to Yugabyte Cloud
Note: Instructions for how to connect this application to YugabyteDB are in
spring-petclinic/src/main/resources/db/yugabytedb/petclinic_db_setup_yugabytedb.md from the repo you clone below.
On your computer from terminal, clone the Spring Boot PetClinic sample application:
git clone https://github.com/yugabyte/spring-petclinic.git.1$ git clone https://github.com/yugabyte/spring-petclinic.git2Cloning into 'spring-petclinic'...3remote: Enumerating objects: 8616, done.4remote: Counting objects: 100% (18/18), done.5remote: Compressing objects: 100% (18/18), done.6remote: Total 8616 (delta 1), reused 13 (delta 0), pack-reused 85987Receiving objects: 100% (8616/8616), 7.29 MiB | 19.03 MiB/s, done.8Resolving deltas: 100% (3268/3268), done.
Copy the contents of
In Yugabyte Cloud, click on “Connect”. Click on Connect to your Application.
Save your host address and port (highlighted in the screenshot below) for later. Click the Back link in the Connect to Cluster navigation breadcrumb.
Click the Launch Cloud Shell button.
Leave the default Database Name, Database User Name, and API Type. Click the Confirm button. This will launch the Cloud Shell.
Enter the admin password you copied previously. After you enter the password, you will have a standard YSQL shell (exactly like a PSQL shell in PostgreSQL) that you can interact with from your browser.1Password for user admin:2ysqlsh (11.2-YB-22.214.171.124-b0)3SSL connection (protocol: TLSv1.2, cipher: ECDHE-RSA-AES256-GCM-SHA384, bits: 256, compression: off)4Type "help" for help.56yugabyte=#
Go to http://localhost:8080. The PetClinic sample application should be available.
The PetClinic sample application is now connected to your Yugabyte Cloud cluster and running locally.
Containerize the Spring Boot PetClinic sample application
Start Docker on your computer and containerize your Spring Boot PetClinic sample application:
Tag your image:
docker tag [image_id] spring-petclinic– you can find your image id by running
docker image ls.
Run your image as a container in Docker to make sure it’s working correctly:
docker run -d --name=spring-petclinic -p 8080:8080 -e JAVA_OPTS="-Dspring.profiles.active=yugabytedb -Dspring.datasource.url=jdbc:postgresql://[host]:[port]/petclinic?load-balance=true -Dspring.datasource.initialization-mode=never" spring-petclinic.
a. Go to http://localhost:8080. The PetClinic sample application should be available.
The PetClinic sample application is now connected to your Yugabyte Cloud cluster and running locally on Docker.
Start Docker (or another minkube-compatible container/VM manager) on your computer.
Open terminal and start minikube:
minikube start1😄 minikube v1.22.0 on Darwin 11.5.1 (arm64)2✨ Automatically selected the docker driver3👍 Starting control plane node minikube in cluster minikube4🚜 Pulling base image ...5💾 Downloading Kubernetes v1.21.2 preload ...6> gcr.io/k8s-minikube/kicbase...: 326.19 MiB / 326.19 MiB 100.00% 10.82 Mi7> preloaded-images-k8s-v11-v1...: 522.45 MiB / 522.45 MiB 100.00% 15.90 Mi8🔥 Creating docker container (CPUs=2, Memory=4000MB) ...9🐳 Preparing Kubernetes v1.21.2 on Docker 20.10.7 ...10▪ Generating certificates and keys ...11▪ Booting up control plane ...12▪ Configuring RBAC rules ...13🔎 Verifying Kubernetes components...14▪ Using image gcr.io/k8s-minikube/storage-provisioner:v515🌟 Enabled addons: storage-provisioner, default-storageclass16🏄 Done! kubectl is now configured to use "minikube" cluster and "default" namespace by default
Open a 2nd terminal and run:
minikube dashboard1🔌 Enabling dashboard ...2▪ Using image kubernetesui/dashboard:v2.1.03▪ Using image kubernetesui/metrics-scraper:v1.0.44🤔 Verifying dashboard health ...5🚀 Launching proxy ...6🤔 Verifying proxy health ...7🎉 Opening http://127.0.0.1:51726/api/v1/namespaces/kubernetes-dashboard/services/http:kubernetes-dashboard:/proxy/ in your default browser...
That will open http://127.0.0.1:51726/api/v1/namespaces/kubernetes-dashboard/services/http:kubernetes-dashboard:/proxy. Keep this tab open. You’ll use it later to check on your application deployment.
Deploy the Spring Boot PetClinic sample application image to minikube
In your 1st terminal, point your shell to minikube’s Docker daemon — This will make it so that Docker and kubectl will use your minikube container registry instead of Docker’s container registry
minikube docker-env.1export DOCKER_TLS_VERIFY="1"2export DOCKER_HOST="tcp://127.0.0.1:58018"3export DOCKER_CERT_PATH="/Users/gavinjohnson/.minikube/certs"4export MINIKUBE_ACTIVE_DOCKERD="minikube"5# To point your shell to minikube's docker-daemon, run:6# eval $(minikube -p minikube docker-env)
eval $(minikube -p minikube docker-env)— last line in output above.
Build your image again:
Tag your image again:
docker tag [image_id] spring-petclinic– you can find your image id by running
docker image ls.
Create a new file named
manifest-minikube.yml, enter the following contents, and save:1apiVersion: v12kind: Service3metadata:4name: spring-petclinic5labels:6 run: spring-petclinic7spec:8selector:9 app: spring-petclinic10ports:11 - protocol: TCP12 port: 808013 targetPort: 808014type: LoadBalancer15---16apiVersion: apps/v117kind: Deployment18metadata:19name: spring-petclinic20labels:21 app: spring-petclinic22spec:23replicas: 224selector:25 matchLabels:26 app: spring-petclinic27template:28 metadata:29 labels:30 app: spring-petclinic31 spec:32 containers:33 - name: spring-petclinic34 image: spring-petclinic35 imagePullPolicy: Never36 ports:37 - containerPort: 808038 env:39 - name: JAVA_OPTS40 value: "-Dspring.profiles.active=yugabytedb -Dspring.datasource.url=jdbc:postgresql://[host]:[port]/petclinic?load-balance=true -Dspring.datasource.initialization-mode=never"
Create the Service and Deployment in minikube:
kubectl create -f manifest-minikube.yml.1$ kubectl create -f manifest-minikube.yml2service/spring-petclinic created3deployment.apps/spring-petclinic created
Open a 3rd terminal and externally expose your load balancer:
minikube tunnel1$ minikube tunnel2🏃 Starting tunnel for service spring-petclinic.
Wait for your application to come online (3 minutes or more for me).
If you go to the local IP and Port you configured in your yaml (http://127.0.0.1:8080), the PetClinic sample application should be available.
The PetClinic sample application is now connected to your Yugabyte Cloud cluster and running on Kubernetes locally on minikube.
If all else fails:
kubectl delete -f manifest-minikube.yml
minikube delete --all