gRPC Health Check Monitor

gRPC monitors call the standard gRPC Health Checking Protocol (grpc.health.v1.Health/Check) and map the response to monitor status.

Minimum setup

Set:

host
port (default 50051)

Optionally set service to check a specific service name instead of overall server health.

Status logic

The ServingStatus from the health check response maps to:

SERVING → UP
NOT_SERVING → DOWN
UNKNOWN / SERVICE_UNKNOWN → DEGRADED

Connection errors and timeouts return DOWN.

Configuration fields

Field	Type	Default	Notes
`host`	`string`	—	Required
`port`	`number`	`50051`	Required
`service`	`string`	`""`	Fully qualified service name; empty = overall
`tls`	`boolean`	`false`	Use TLS credentials
`insecure`	`boolean`	`false`	Skip TLS certificate verification (TLS only)
`timeout`	`number`	`10000`	Request deadline in ms

Example

{
    "type": "GRPC",
    "type_data": {
        "host": "grpc.example.com",
        "port": 50051,
        "service": "my.package.MyService",
        "tls": true,
        "timeout": 5000
    }
}

Self-signed TLS example

{
    "type": "GRPC",
    "type_data": {
        "host": "grpc.example.com",
        "port": 50051,
        "tls": true,
        "insecure": true,
        "timeout": 5000
    }
}

Troubleshooting

Immediate DOWN: wrong host/port, service not running, or firewall blocking the connection
DEGRADED: server returned UNKNOWN or SERVICE_UNKNOWN — verify the service name is registered
Timeout: increase timeout or check network latency to the gRPC server
TLS errors: ensure the server has a valid certificate, disable insecure, or check if TLS should be disabled
Self-signed certificate (messages mentioning --use-system-ca): set tls: true and insecure: true on the monitor (same as grpcurl -insecure). Alternatively, trust the issuer: run Node with NODE_OPTIONS='--use-system-ca' so the OS CA store is used (Node 22+ defaults to bundled CA roots only)