On how many hosts did the service run and how many instances were launched?

To determine how many hosts a service was running on and how many instances were launched, monitoring and orchestration systems such as Kubernetes, Docker Swarm, or custom scripts are usually used.

For example: if the service is deployed in Kubernetes, you can run the command:

kubectl get pods -l app=your-service -o wide

This will show a list of pods (instances) and the hosts (nodes) on which they are running.

In Go code, you can integrate metrics and logging to track the number of active instances and their distribution across hosts. For example, using Prometheus and exporting metrics with a client:

import (
    "github.com/prometheus/client_golang/prometheus"
)

var (
    instanceCount = prometheus.NewGaugeVec(
        prometheus.GaugeOpts{
            Name: "service_instance_count",
            Help: "Number of service instances per host",
        },
        []string{"host"},
    )
)

func init() {
    prometheus.MustRegister(instanceCount)
}

func updateInstanceCount(host string, count int) {
    instanceCount.WithLabelValues(host).Set(float64(count))
}

Thus, you can programmatically and through monitoring tools find out the number of hosts and instances.