Helidon: Microservice Health Check

a health check of our microservice is part of microprofile specification. we don't need to write anything special with helidon. all the health api will be available out of the box in our implementation.

health check apis

there are three endpoints are available as part of the specification to check the health of our microservices. these apis will be used by the 3rd party application/integration tools to check the health. those apis are as follows,

/health /health/live /health/ready

you may also like: dockerized app with java module system, helidon, and alpine-based jdk 12 ea

/health

this is the top-level api to check the health of our microservice. apart from the liveness and readiness of our application, we will also get more internal information about services like the following,

/health/live

we can check the liveness of our application using this end-point. any micro profile compliant framework should offer this implementation out of the box. helidon will also offer such end-point and return meaningful information. in some cases, we may need to customize the liveness end-points.

in such use-cases, we need to provide an implementation of the healthcheck interface with @liveness annotation. see the code below.

//custom liveness implementation @liveness @applicationscoped 
//public class dailyapiliveness implements healthcheck { 
//@override public healthcheckresponse call() { return 
//healthcheckresponse.named("dailyapi liveness") .up() 
//.withdata("time", system.currenttimemillis()) .build(); } }

/health/ready

this end-point is similar to the above live end-point. the only difference is the @readiness annotation instead of @liveness. we can add our custom logic to calculate the readiness to expose the real value to our partners or integration solutions. please see the sample code below,

//we can add our custom logic to calculate the real readiness and 
//return the response //accordingly. @readiness @applicationscoped 
//public class dailyapireadiness implements healthcheck { 
//@override public healthcheckresponse call() { return 
//healthcheckresponse.named("dailyapi readiness") .up() .build(); } }

once everything is set properly, we will get the following response when we access the /health api.

{ "outcome": "up", "status": "up", "checks": [ { "name": "dailyapi liveness", 
  "state": "up", "status": "up", "data": { "time": 1571677530451 } }, 
{ "name": "dailyapi readiness", "state": "up", "status": "up", "data": 
 { "time": 1571677530460 } }, { "name": "deadlock", "state": "up", "status": 
  "up" }, { "name": "diskspace", "state": "up", "status": "up", "data": 
  { "free": "96.31 gb", "freebytes": 103412359168, "percentfree": 
       "41.23%", "total": "233.57 gb", "totalbytes": 250790436864 } }, 
   { "name": "heapmemory", "state": "up", "status": "up", "data": 
    { "free": "194.89 mb", "freebytes": 204357016, "max": "1.78 gb", 
     "maxbytes": 1908932608, "percentfree": "98.02%", "total": "231.00 mb", 
     "totalbytes": 242221056 } } ] }

i hope you have enjoyed the article. let me know if you need any improvement or suggestion on this topic.

happy coding!

further reading

microprofile: what you need to know

choosing a rest framework for microservices

thoughts on quarkus