The Central Go Modules Repository

To use GoCenter:
export GOPROXY=
March 31st 2020
Last Modified


A CloudStack API client enabling Go programs to interact with CloudStack in a simple and uniform way


This package covers the complete CloudStack API and is well tested. Of course there will still be untested corner cases when you have over 400 API commands that you can use, but over all it’s safe to use this package.

To be able to find the API command you want, they are grouped by ‘services’ which match the grouping you can see/find on the CloudStack API docs website.


The cloudstack package is always generated against the latest stable CloudStack release (currently v4.11.x). Luckily the API doesn’t change that much, and were it does we try to make sure the generated package is able handle both the old and the new case. Over time it will be impossible to support all version with just one package, but until now we seem to manage this pretty well.

Please see the package documentation on GoDocs.


// Create a new API client
cs := cloudstack.NewAsyncClient("", "your-api-key", "your-api-secret", false)

// Create a new parameter struct
p := cs.VirtualMachine.NewDeployVirtualMachineParams("service-offering-id", "template-id", "zone-id")

// Set the name
name := "server-1"

// Set the display name
p.SetDisplayname("Test server 1")

// Set any other options required by your setup in the same way

// Create the new instance
r, err := cs.VirtualMachine.DeployVirtualMachine(p)
if err != nil {
	log.Fatalf("Error creating the new instance %s: %s", name, err)

fmt.Printf("UUID or the newly created machine: %s", r.ID)


Next to the API commands CloudStack itself offers, there are a few additional features/function that are helpful. For starters there are two clients, an normal one (created with NewClient(...)) and an async client (created with NewAsyncClient(...)). The async client has a buildin waiting/polling feature that waits for a configured amount of time (defaults to 300 seconds) on running async jobs. This is very helpfull if you do not want to continue with your program execution until the async job is done.

There is also a function you can call manually (GetAsyncJobResult(...)) that does the same, but then as a seperate call after you started the async job.

Another nice feature is the fact that for every API command you can create the needed parameter struct using a New...Params function, like for example NewListTemplatesParams. The advantage of using this functions to create a new parameter struct, is that these functions know what the required parameters are of ever API command, and they require you to supply these when creating the new struct. Every additional paramater can be set after creating the struct by using SetName() like functions.

Last but not least there are a whole lot of helper function that will try to automatically find an UUID for you for a certain item (disk, template, virtualmachine, network…). This makes it much easier and faster to work with the API commands and in most cases you can just use then if you know the name instead of the UUID.


I fully understand I need to document this all a little more/better and there should also be some tests added.

Getting Help

Please try to see if GoDocs can provide some answers first!


Sander van Harmelen (


Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at