srvmon

Service health monitoring for Go microservices. gRPC + REST, Kubernetes-ready.

Install

# Library
go get github.com/s4bb4t/srvmon

# CLI
go install github.com/s4bb4t/srvmon/cmd/srvmon-cli@latest

Quick Start

cfg := srvmon.Config{
    Version:     "1.0.0",
    GRPCAddress: ":50051",
    HTTPAddress: ":8080",
}

monitor := srvmon.New(cfg, logger)
monitor.AddDependencies(
    NewPingChecker("redis", "localhost:6379", 5*time.Second, true),       // critical
    srvmon.NewConnChecker(grpcConn, "auth-svc", true),                    // gRPC health check
    NewPingChecker("metrics", "localhost:9090", 2*time.Second, false),    // non-critical
)
monitor.SetReady()
monitor.Run(ctx) // blocks until ctx is canceled

How It Works

Implement the Checker interface and register dependencies:

type Checker interface {
    MustOK(ctx context.Context) bool                       // true = critical dependency
    Check(ctx context.Context) (*pb.CheckResult, error)    // perform the check
}

Status aggregation:

Dependency fails	MustOK() = true	MustOK() = false
Result	DOWN	DEGRADED

If all checks pass, service status is UP.

Built-in: ConnChecker

Verifies gRPC dependencies via the standard grpc.health.v1.Health/Check protocol — not just connection state, but actual service readiness.

conn, _ := grpc.NewClient("localhost:50052", grpc.WithTransportCredentials(insecure.NewCredentials()))

// Basic
srvmon.NewConnChecker(conn, "auth", true)

// With options
srvmon.NewConnChecker(conn, "auth", true,
    srvmon.WithTimeout(2*time.Second),
    srvmon.WithService("my.service.Name"),  // check specific service
)

The target server must register grpc.health.v1.Health (srvmon does this automatically for its own gRPC server).

Endpoints

HTTP	gRPC	Description
GET /health	srvmon.v1.srvmon/Health	Full health report
GET /healthz	—	Alias for /health
GET /ready	srvmon.v1.srvmon/Ready	Readiness probe
GET /readyz	—	Alias for /ready

srvmon also registers grpc.health.v1.Health on its gRPC server, so ConnChecker from other services works out of the box.

CLI

  srvmon — service health monitor
  ────────────────────────────────────────────────

  HEALTH  Health:  UP   v1.0.0

  ├── redis         ● UP      connection successful
  ├── auth-svc      ● UP      SERVING
  └── metrics       ● DOWN    connection failed
        dial tcp: connection refused

  READY  Readiness:  NOT READY   connection failed

srvmon-cli                              # localhost:8080
srvmon-cli -a localhost:9090            # custom address
srvmon-cli --watch                      # live updates (in-place, no spam)
srvmon-cli -w -i 1s -a localhost:8085   # watch with 1s interval
srvmon-cli health                       # health only
srvmon-cli ready                        # readiness only

Flag	Short	Default	Description
--addr	-a	localhost:8080	HTTP address
--timeout	-t	3s	Request timeout
--watch	-w	false	Poll and update in-place
--interval	-i	2s	Poll interval

Kubernetes

livenessProbe:
  httpGet:
    path: /healthz
    port: 8080
readinessProbe:
  httpGet:
    path: /readyz
    port: 8080

Or native gRPC probes (k8s 1.24+):

livenessProbe:
  grpc:
    port: 50051

License

MIT