How to Manage Rackspace Cloud Servers
June 22, 2011
A new ebook to help users of Rackspace setup and manage their cloud-based servers is out this week. Called Managing Rackspace Cloud Servers and written by Jacek Artymiak, it is a very practical and useful guide, and seems to be unique.
The 75 page book teaches you how to use the openstack-compute Python-based tool to manage server farms built on top of the Rackspace Cloud Servers service. The author takes you through the steps to get the source code installed on your machines and illustrates some basic management tasks with copious examples and some geek humor thrown in for extra measure. Even if you aren’t familiar with Python you should be comfortable using this tool.
For example, you can use openstack-compute to resize RAM or disk on one of your VMs, rebuild a VM from a known working version, and making backups of your cloud servers.
Rackspace’s programming interfaces make automating common tasks possible, and why these Python-based tools can be useful. "Consider the following scenario: you have an nginx front-end server as a load balancer and a couple of Tornado web servers behind it servicing the incoming traffic. Suppose you get famous and the load on your web servers rises above a certain threshold," says Artymiak…
"What you want to do is create a new Tornado web server instance, add the new server’s IP address to the nginx configuration file and restart nginx. Then you will have three web servers sharing the load. If you monitor the load on the web servers using Nagios, you can configure it to run your custom scripts when certain events happen. In our case that event would be the rise in the web server load. Before cloud computing APIs, your Nagios monitoring box would send you an email and you would have to log in and create a new web server manually.
Nowadays, thanks to APIs and the tools that wrap them in a convenient form, like openstack-compute does, you can write scripts that do all of that work for you. If you are clever (or lazy…), you can also write scripts that delete instances no longer needed when the server load drops. Your Nagios monitoring box can now send you an email describing what it’s done while you were sleeping."