Configure Amazon ECS HealthCheck using paketo-buildpacks/health-checker

[mgdvinuesa-knowmadmood]

Hello, I have a spring boot application that uses paketo generate and deploy the OCI image over AWS ECS Service.
My problem is that I can't configure the health check to use the paketo health-checker.

I build the image and it works perfectly

mvn spring-boot:build-image  \
  -Dspring-boot.build-image.imageName=XXXXXXXX \
  -s /settings.xml

Log (the important part):

[INFO]     [creator]     Paketo Buildpack for Health Checkers 2.9.0
[INFO]     [creator]       https://github.com/paketo-buildpacks/health-checker
[INFO]     [creator]       Build Configuration:
[INFO]     [creator]         $BP_HEALTH_CHECKER_DEPENDENCY  thc   which health checker to contribute
[INFO]     [creator]         $BP_HEALTH_CHECKER_ENABLED     true  contributes a health checker if enabled
[INFO]     [creator]       Tiny Health Checker 0.37.0: Contributing to layer
[INFO]     [creator]         Downloading from https://github.com/dmikusa/tiny-health-checker/releases/download/v0.37.0/tiny-health-checker-x86_64-unknown-linux-musl.tar.xz
[INFO]     [creator]         Verifying checksum
[INFO]     [creator]         Expanding to /layers/paketo-buildpacks_health-checker/thc
[INFO]     [creator]         The Tiny Health Checker binary is available at /layers/paketo-buildpacks_health-checker/thc/bin/thc
[INFO]     [creator]         A symlink is available at /workspace/health-check for your convenience

In my local I run the image and the health works:

docker run --env-file ../.env  <image> -p 8080:8080
docker ps
docker exec -it fee146ab7cf8 /workspace/health-check

But when I configure the AWS ECS HealthCheck, it does not work

 "healthCheck": {
                "command": [
                  "CMD",
                  "/workspace/health-check || exit 1"
                ],
                "interval": 45,
                "timeout": 5,
                "retries": 6,
                "startPeriod": 60
            },

Using CMD as the interface to run the command is not valid ?

More information:
https://docs.aws.amazon.com/AmazonECS/latest/developerguide/healthcheck.html
https://docs.aws.amazon.com/AmazonECS/latest/developerguide/task_definition_parameters.html#container_definition_healthcheck

thanks in advance!

[dmikusa]

Have you considered just using the ALB HTTP health checks here?

I run a number of services on ECS and since Spring Boot Actuators provides a /actuators/health endpoint, doing health checks is just a matter of telling my ALB to use /actuator/health as the health check endpoint for the container. Those requests come into the app, if one fails, it will end up doing the same thing and causing the task to be recreated.

Sorry, I know that doesn't necessarily answer your question. I haven't tried running the health checker in the container, but what you're doing looks reasonable. Do you get anything in your logs? Or any further output? If the health-check process fails, it will write the problem to STDOUT.

[herder]

Hi, have you added the THC_PATH and THC_PORT variables to the task definition?
Previously you could bake them into the image, but now you need to set them when running the container.