- Liveness probe available at `/minio/health/live`
- Cluster probe available at `/minio/health/cluster`

Read more on how to use these endpoints in [MinIO healthcheck guide](https://github.com/minio/minio/blob/master/docs/metrics/healthcheck/README.md).

## Prometheus Probe

```sh
mc admin trace --verbose myminio
```

To trace entire HTTP request and also internode communication

```sh
mc admin trace --all --verbose myminio
```

## Subnet Health

quarkus/extensions/smallrye-health/deployment/pom.xml
	quarkus/core/deployment/pom.xml
	quarkus/extensions/arc/deployment/pom.xml
	quarkus/extensions/vertx-http/deployment/pom.xml
	quarkus/extensions/jsonp/deployment/pom.xml
	quarkus/extensions/smallrye-health/spi/pom.xml
	quarkus/extensions/smallrye-openapi/spi/pom.xml
	quarkus/extensions/smallrye-health/runtime/pom.xml
	quarkus/test-framework/junit5-internal/pom.xml

import org.gradle.work.DisableCachingByDefault
import java.io.File

@DisableCachingByDefault
abstract class FixProjectHealthTask : DefaultTask() {

    private val errorLogFile = File("gradle-project-health.log")

    @TaskAction
    fun fixProjectHealth() {
        if (!errorLogFile.exists()) {
            println("Error log not found! Run the projectHealth check first and redirect output.")
            return
        }

- Kube-proxy extends the schema of its healthz/ and livez/ endpoints to incorporate information about the corresponding IP family ([#129271](https://github.com/kubernetes/kubernetes/pull/129271), [@aroradaman](https://github.com/aroradaman)) [SIG Network and Windows]

		Online:   true,
	}
	sys.hMutex.Unlock()
}

// newHCClient initializes an anonymous client for performing health check on the remote endpoints
func newHCClient() *madmin.AnonymousClient {
	clnt, e := madmin.NewAnonymousClientNoEndpoint()
	if e != nil {
		bugLogIf(GlobalContext, errors.New("Unable to initialize health check client"))
		return nil
	}
	clnt.SetCustomTransport(globalRemoteTargetTransport)
	return clnt
}

- The slow initialization of container runtime will not cause System WatchDog to kill kubelet. Device Manager is not considered healthy before it attempted to start listening on the port. ([#135207](https://github.com/kubernetes/kubernetes/pull/135207), [@SergeyKanzhelev](https://github.com/SergeyKanzhelev)) [SIG Node]

## Dependencies

### Added

	storageRESTVersionPrefix = SlashSeparator + storageRESTVersion
	storageRESTPrefix        = minioReservedBucketPath + "/storage"
)

const (
	storageRESTMethodHealth = "/health"

	storageRESTMethodAppendFile     = "/afile"
	storageRESTMethodCreateFile     = "/cfile"
	storageRESTMethodWriteAll       = "/wall"
	storageRESTMethodReadVersion    = "/rver"
	storageRESTMethodReadXL         = "/rxl"

    //
    private static final Logger logger = LogManager.getLogger(AdminMaintenanceAction.class);

    private static final String[] CAT_NAMES =
            { "aliases", "allocation", "count", "fielddata", "health", "indices", "master", "nodeattrs", "nodes", "pending_tasks",
                    "plugins", "recovery", "repositories", "thread_pool", "shards", "segments", "snapshots", "templates" };

	Type        string `json:"type"`
	Description string `json:"description"`
	Optional    bool   `json:"optional"`

	// Indicates if the value contains sensitive info that shouldn't be exposed
	// in certain apis (such as Health Diagnostics/Callhome)
	Sensitive bool `json:"-"`

	// Indicates if the value is a secret such as a password that shouldn't be
	// exposed by the server
	Secret bool `json:"-"`