Service Discovery (Part 2): Java, a Database App, and Kubernetes

This article will show how a simple Java application can talk to a database using service discovery in Kubernetes. Service Discovery with Java and Database application in DC/OS explains why service discovery is an important aspect for a multi-container application. That blog also explained how this can be done for DC/OS.

Let’s see how this can be accomplished in Kubernetes with a single instance of an application server and a database server. This blog will use WildFly for the application server and Couchbase for the database.

This blog will use the following main steps:

• Start Kubernetes one-node cluster
• Kubernetes application definition
• Deploy the application
• Access the application

Start Kubernetes Cluster

Minikube is the easiest way to start a one-node Kubernetes cluster in a VM on your laptop. The binary needs to be downloaded first and then installed.

Complete installation instructions are available at github.com/kubernetes/minikube.

The latest release can be installed on OSX as:

curl -Lo minikube https://storage.googleapis.com/minikube/releases/v0.17.1/minikube-darwin-amd64 \
&& chmod +x minikube

It also requires kubectl to be installed. The article Installing and Setting up kubectl provides detailed instructions on how to setup kubectl. On OSX, it can be installed as:

curl -LO https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/darwin/amd64/kubectl \
  && chmod +x ./kubectl

Now, start the cluster as:

minikube start
Starting local Kubernetes cluster...
Starting VM...
Downloading Minikube ISO
 88.71 MB / 88.71 MB [==============================================] 100.00% 0s
SSH-ing files into VM...
Setting up certs...
Starting cluster components...
Connecting to cluster...
Setting up kubeconfig...
Kubectl is now configured to use the cluster.

The kubectl version command shows more details about the kubectl client and minikube server version:

kubectl version
Client Version: version.Info{Major:"1", Minor:"5", GitVersion:"v1.5.4", GitCommit:"7243c69eb523aa4377bce883e7c0dd76b84709a1", GitTreeState:"clean", BuildDate:"2017-03-07T23:53:09Z", GoVersion:"go1.7.4", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"5", GitVersion:"v1.5.3", GitCommit:"029c3a408176b55c30846f0faedf56aae5992e9b", GitTreeState:"clean", BuildDate:"1970-01-01T00:00:00Z", GoVersion:"go1.7.3", Compiler:"gc", Platform:"linux/amd64"}

More details about the cluster can be obtained using the kubectl cluster-infocommand:

Kubernetes master is running at https://192.168.99.100:8443
KubeDNS is running at https://192.168.99.100:8443/api/v1/proxy/namespaces/kube-system/services/kube-dns
kubernetes-dashboard is running at https://192.168.99.100:8443/api/v1/proxy/namespaces/kube-system/services/kubernetes-dashboard

To further debug and diagnose cluster problems, use 'kubectl cluster-info dump'.

Kubernetes Application Definition

The application definition is defined at github.com/arun-gupta/kubernetes-java-sample/blob/master/service-discovery.yml. It consists of:

• A Couchbase service
• Couchbase replica set with a single pod
• A WildFly replica set with a single pod

apiVersion: v1
kind: Service
metadata: 
  name: couchbase-service
spec: 
  selector: 
    app: couchbase-rs-pod
  ports:
    - name: admin
      port: 8091
    - name: views
      port: 8092
    - name: query
      port: 8093
    - name: memcached
      port: 11210
---
apiVersion: extensions/v1beta1
kind: ReplicaSet
metadata:
  name: couchbase-rs
spec:
  replicas: 1
  template:
    metadata:
      labels:
        app: couchbase-rs-pod
    spec:
      containers:
      - name: couchbase
        image: arungupta/couchbase:travel
        ports:
        - containerPort: 8091
        - containerPort: 8092
        - containerPort: 8093
        - containerPort: 11210
---
apiVersion: extensions/v1beta1
kind: ReplicaSet
metadata:
  name: wildfly-rs
  labels:
    name: wildfly
spec:
  replicas: 1
  template:
    metadata:
      labels:
        name: wildfly
    spec:
      containers:
      - name: wildfly-rs-pod
        image: arungupta/wildfly-couchbase-javaee:travel
        env:
        - name: COUCHBASE_URI
          value: couchbase-service
        ports:
        - containerPort: 8080

The key part is where the value of the COUCHBASE_URI environment variable is the name of the Couchbase service. This allows the application deployed in WildFly to dynamically discover the service and communicate with the database.

The arungupta/couchbase:travel Docker image is created using github.com/arun-gupta/couchbase-javaee/blob/master/couchbase/Dockerfile.

The arungupta/wildfly-couchbase-javaee:travel Docker image is created using github.com/arun-gupta/couchbase-javaee/blob/master/Dockerfile.

The Java EE application waits for database initialization to be complete before it starts querying the database. This can be seen at github.com/arun-gupta/couchbase-javaee/blob/master/src/main/java/org/couchbase/sample/javaee/Database.java#L25.

Deploy Application

This application can be deployed as:

kubectl create -f ~/workspaces/kubernetes-java-sample/service-discovery.yml

The list of services and replica set can be shown using the command kubectl get svc,rs:

NAME                    CLUSTER-IP   EXTERNAL-IP   PORT(S)                                AGE
svc/couchbase-service   10.0.0.97    <none>        8091/TCP,8092/TCP,8093/TCP,11210/TCP   27m
svc/kubernetes          10.0.0.1     <none>        443/TCP                                1h
svc/wildfly-rs          10.0.0.252   <none>        8080/TCP                               21m

NAME              DESIRED   CURRENT   READY     AGE
rs/couchbase-rs   1         1         1         27m
rs/wildfly-rs     1         1         1         27m

Logs for the single replica of Couchbase can be obtained using the command kubectl logs rs/couchbase-rs:

++ set -m
++ sleep 25
++ /entrypoint.sh couchbase-server
Starting Couchbase Server -- Web UI available at http://<ip>:8091 and logs available in /opt/couchbase/var/lib/couchbase/logs
++ curl -v -X POST http://127.0.0.1:8091/pools/default -d memoryQuota=300 -d indexMemoryQuota=300

. . .

{"storageMode":"memory_optimized","indexerThreads":0,"memorySnapshotInterval":200,"stableSnapshotInterval":5000,"maxRollbackPoints":5,"logLevel":"info"}[]Type: 
++ echo 'Type: '
++ '[' '' = WORKER ']'
++ fg 1
/entrypoint.sh couchbase-server

Logs for the WildFly replica set can be seen using the command kubectl logs rs/wildfly-rs:

=========================================================================

  JBoss Bootstrap Environment

  JBOSS_HOME: /opt/jboss/wildfly

. . .

06:32:08,537 INFO  [com.couchbase.client.core.node.Node] (cb-io-1-1) Connected to Node couchbase-service
06:32:09,262 INFO  [com.couchbase.client.core.config.ConfigurationProvider] (cb-computations-3) Opened bucket travel-sample
06:32:09,366 INFO  [stdout] (ServerService Thread Pool -- 65) Sleeping for 3 secs ...
06:32:12,369 INFO  [stdout] (ServerService Thread Pool -- 65) Bucket found!
06:32:14,194 INFO  [org.jboss.resteasy.resteasy_jaxrs.i18n] (ServerService Thread Pool -- 65) RESTEASY002225: Deploying javax.ws.rs.core.Application: class org.couchbase.sample.javaee.MyApplication
06:32:14,195 INFO  [org.jboss.resteasy.resteasy_jaxrs.i18n] (ServerService Thread Pool -- 65) RESTEASY002200: Adding class resource org.couchbase.sample.javaee.AirlineResource from Application class org.couchbase.sample.javaee.MyApplication
06:32:14,310 INFO  [org.wildfly.extension.undertow] (ServerService Thread Pool -- 65) WFLYUT0021: Registered web context: /airlines
06:32:14,376 INFO  [org.jboss.as.server] (ServerService Thread Pool -- 34) WFLYSRV0010: Deployed "airlines.war" (runtime-name : "airlines.war")
06:32:14,704 INFO  [org.jboss.as] (Controller Boot Thread) WFLYSRV0060: Http management interface listening on http://127.0.0.1:9990/management
06:32:14,704 INFO  [org.jboss.as] (Controller Boot Thread) WFLYSRV0051: Admin console listening on http://127.0.0.1:9990
06:32:14,705 INFO  [org.jboss.as] (Controller Boot Thread) WFLYSRV0025: WildFly Full 10.1.0.Final (WildFly Core 2.2.0.Final) started in 29470ms - Started 443 of 691 services (404 services are lazy, passive or on-demand)

Access Application

The kubectl proxy command starts a proxy to the Kubernetes API server. Let’s start a Kubernetes proxy to access our application:

kubectl proxy
Starting to serve on 127.0.0.1:8001

Expose the WildFly replica set as a service using:

kubectl expose --name=wildfly-service rs/wildfly-rs

The list of services can be seen again using kubectl get svc command:

kubectl get svc
NAME                CLUSTER-IP   EXTERNAL-IP   PORT(S)                                AGE
couchbase-service   10.0.0.97    <none>        8091/TCP,8092/TCP,8093/TCP,11210/TCP   41m
kubernetes          10.0.0.1     <none>        443/TCP                                1h
wildfly-service     10.0.0.169   <none>        8080/TCP                               5s

Now, the application is accessible at:

curl http://localhost:8001/api/v1/proxy/namespaces/default/services/wildfly-service/airlines/resources/airline

A formatted output looks like:

[
  {
    "travel-sample": {
      "country": "United States",
      "iata": "Q5",
      "callsign": "MILE-AIR",
      "name": "40-Mile Air",
      "icao": "MLA",
      "id": 10,
      "type": "airline"
    }
  },
  {
    "travel-sample": {
      "country": "United States",
      "iata": "TQ",

. . .

     "name": "Airlinair",
      "icao": "RLA",
      "id": 1203,
      "type": "airline"
    }
  }
]

Now, new pods may be added as part of Couchbase service by scaling the replica set. Existing pods may be terminated or get rescheduled. But the Java EE application will continue to access the database service using the logical name.

This blog showed how a simple Java application can talk to a database using service discovery in Kubernetes.

For further information check out:

• Kubernetes Docs
• Couchbase on Containers
• Couchbase Developer Portal
• Ask questions on Couchbase Forums or Stack Overflow
• Download Couchbase