Define custom health checks
While you coded the database configuration in the previous exercise, Quarkus has generated Kubernetes and OpenShift resources in target/kubernetes
:
target/kubernetes
|-- kubernetes.json
|-- kubernetes.yml
|-- openshift.json
`-- openshift.yml
Inspect their content and observe that you already have a base for your Kubernetes deployment. But before doing that, let’s implement custom health checks.
Add the health monitoring extension
In the terminal window please execute the following command:
./mvnw quarkus:add-extension -Dextensions="io.quarkus:quarkus-smallrye-health"
If you did not stop DevMode and inspect again target/kubernetes/kubernetes.yml
or target/kubernetes/openshift.yml
will notice that it had changed. Your Deployment/DeploymentConfig resource will contain a few lines like to:
livenessProbe:
failureThreshold: 3
httpGet:
path: /q/health/live
port: 8080
scheme: HTTP
initialDelaySeconds: 0
periodSeconds: 30
successThreshold: 1
timeoutSeconds: 10
name: tutorial-app
ports:
- containerPort: 8080
name: http
protocol: TCP
readinessProbe:
failureThreshold: 3
httpGet:
path: /q/health/ready
port: 8080
scheme: HTTP
initialDelaySeconds: 0
periodSeconds: 30
successThreshold: 1
timeoutSeconds: 10
You can access these endpoints also in your local at http://localhost:8080/q/health/ready or http://localhost:8080/q/health/live. |
Default health checks
In dev mode, all your heath checks are visible in health UI: http://localhost:8080/q/health-ui/.
Some extensions may provide default health checks, including that the extension will automatically register its health checks.
For example, quarkus-agroal
(that is used to manage Quarkus datasources) automatically registers a readiness health check that will validate each datasource.
Quarkus has automatic readiness probes added when you use certain extensions:
-
datasource A probe to check database connection status.
-
kafka A probe to check kafka connection status. In this case you need to enable manually by setting quarkus.kafka.health.enabled to true.
-
mongoDB A probe to check MongoDB connection status.
-
neo4j A probe to check Neo4J connection status.
-
artemis A probe to check Artemis JMS connection status.
-
kafka-streams Liveness (for stream state) and Readiness (topics created) probes.
-
vault A probe to check Vault conection status.
-
gRPC A readiness probe for the gRPC services.
-
Cassandra A readiness probe to check Cassandra connection status.
-
Redis A readiness probe to check Redis connection status.
Customize integrations and startup configurations
Let’s assume that at application startup we insert some data in our database, but we should also update the message content from an external service. This type of scenario requires customizing health checks for the readiness probe to take into account the second dependency.
Add the REST Client extension
In the terminal window please execute the following command :
./mvnw quarkus:add-extension -Dextensions="io.quarkus:quarkus-rest-client"
Consuming a REST endpoint
Implement a simple data access object (DTO) to help reading the data:
package com.redhat.developers;
import com.fasterxml.jackson.annotation.JsonIgnoreProperties;
@JsonIgnoreProperties(ignoreUnknown = true)
public class ExternalGreeting {
public String hello;
}
And now let’s define an interface and register it as a REST client:
package com.redhat.developers;
import org.eclipse.microprofile.rest.client.inject.RegisterRestClient;
import jakarta.ws.rs.GET;
import jakarta.ws.rs.Path;
import jakarta.ws.rs.Produces;
import jakarta.ws.rs.QueryParam;
import jakarta.ws.rs.core.MediaType;
@RegisterRestClient
@Path("/hellosalut")
public interface HelloService {
@GET
@Path("/")
@Produces(MediaType.APPLICATION_JSON)
ExternalGreeting getContent(@QueryParam("lang") String lang);
}
And add in src/main/resources/application.properties
:
com.redhat.developers.HelloService/mp-rest/url=https://fourtonfish.com
Define actions at startup initialization
Furthermore, we can use the previously defined endpoint to update our data using an intermediary repository:
package com.redhat.developers;
import io.quarkus.hibernate.orm.panache.PanacheRepository;
import io.quarkus.panache.common.Parameters;
import jakarta.enterprise.context.ApplicationScoped;
import jakarta.transaction.Transactional;
@ApplicationScoped
public class GreetingRepository implements PanacheRepository<Message> {
@Transactional
public int update(String content, String language) {
return update("content= :content where language= :language ",
Parameters.with("content", content)
.and("language", language));
}
}
And inject this class in the one used to customize startup initialization of data:
package com.redhat.developers;
import io.quarkus.arc.profile.UnlessBuildProfile;
import io.quarkus.runtime.Startup;
import jakarta.annotation.PostConstruct;
import jakarta.enterprise.context.ApplicationScoped;
import jakarta.inject.Inject;
import org.eclipse.microprofile.rest.client.inject.RestClient;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.List;
@Startup (1)
@ApplicationScoped
@UnlessBuildProfile("test") (2)
public class MessageInitializer {
private static final Logger LOGGER = LoggerFactory.getLogger(MessageInitializer.class);
@Inject
@RestClient
HelloService helloService; (3)
@Inject
GreetingRepository repository; (4)
@PostConstruct
public void init() {
LOGGER.debug("Updating the db from external service");
List<Message> messages = Message.findAll().list();
for (Message message : messages) {
String language = message.getLanguage();
repository.update(helloService.getContent(language).hello, language);
}
LOGGER.debug("End update of the db ");
}
}
1 | This annotation initializes a CDI bean at application startup.
This code will be executed after initializing the database from import.sql . |
2 | Enable for both prod and dev build time profiles. |
3 | Inject the RestClient service. |
4 | Inject the service that updates database content and has @Transactional annotation set on the invoked method. |
5 | Invoke record update. |
Customize health endpoints and readiness probe
You can change the root path to the health endpoints by setting the following property in src/main/resources/application.properties
:
quarkus.smallrye-health.root-path=/health
The Quarkus Kubernetes/OpenShift extension will take into account your custom probe definitions when generating their YAML.
If you reload the context in DevMode (by pressing s
), you would notice that your Kubernetes/OpenShift manifests have changed and take into account your new configuration.
As the database readiness is already assessed, we can customize another readiness probe to check the availability of the endpoint https://fourtonfish.com
:
package com.redhat.developers;
import io.smallrye.health.checks.UrlHealthCheck;
import jakarta.enterprise.context.ApplicationScoped;
import jakarta.ws.rs.HttpMethod;
import org.eclipse.microprofile.config.inject.ConfigProperty;
import org.eclipse.microprofile.health.HealthCheck;
import org.eclipse.microprofile.health.Readiness;
@ApplicationScoped
public class CustomHealthCheck {
@ConfigProperty(name = "com.redhat.developers.HelloService/mp-rest/url")
String externalURL;
@Readiness (1)
HealthCheck checkURL() {
return new UrlHealthCheck(externalURL+"/hellosalut/?lang=en") (2)
.name("external-url-check").requestMethod(HttpMethod.GET).statusCode(200);
}
}
1 | Annotate the method with org.eclipse.microprofile.health.Readiness to signal its implementation. |
2 | UrlHealthCheck checks if host is reachable using a Http URL connection. |
Quarkus comes with some HealthCheck implementations for you to check status of different components:
|