nadeemc / clouddetect

Go package for resolving IP addresses back to published cloud provider public ranges

clouddetect

Go package to determine if an IP address resolves to one of the major cloud providers. Helpful for risk-scoring potential bot traffic in conjunction with other signals such as User Agent.

Currently the library consumes the published IP ranges from Amazon, Google, and Microsoft.

API usage

package main

import (
	"fmt"
	"os"

	"github.com/99designs/clouddetect"
)

func main() {
    ip := net.ParseIP("127.0.0.1")
    if cloud, err := clouddetect.Resolve(ip); err == nil {
        fmt.Println(cloud.ProviderName)
    }
}

The default Client has an internal cache with a TTL of 12 hours. The first request to resolve an IP will be slow as it fetches all published ranges. See the clouddetect godocs for more detail on the API.

CLI usage

go get github.com/99designs/clouddetect/cli/clouddetect

then

clouddetect -ip=127.0.0.1

Caching to Disk

package main

import (
    "fmt"
    "os"

    "github.com/99designs/clouddetect"
    "time"
)

var (
    client = clouddetect.NewClient(24 * time.Hour)
)

func init() {
    client.CacheFilePath = path.Join(os.TempDir(), "cloud-ip-cache.json")
    // Cache will be persisted and loaded from disk. A lock file is also generated during this process to allow multiple instances to share the same cache file path without having competing refreshes.
    client.RefreshCache()
}

func main() {
    ip := net.ParseIP("127.0.0.1")
    if cloud, err := client.Resolve(ip); err == nil {
        fmt.Println(cloud.ProviderName)
    }
}

Using a CacheFilePath speeds up the initial launch time, so your first request to resolve an IP will only be slow if no local cache exists. Optionally, manually calling the client.RefreshCache() function in your init() ensures your cache is ready to go even for the first request.

LICENSE

MIT 2018 99designs