Explaining the System Disk and Data Disk Architecture
Performance Cloud Servers for Linux and Microsoft Windows use a new architecture that pairs a system disk with a separate data disk(s). The benefits of implementing this new architecture include improved provisioning times, as well as more flexibility for data storage and disaster recovery. Snapshots or images of Performance servers will only save the information from the system disk consisting of your operating system and configurations. This means that you will need to use a separate method for saving information located on the data disk(s). Because images now preserve only the system disk information, imaging will complete more quickly and provisioning new servers from a saved system disk image will be faster.
Saving Your Configuration Using System Images
Each Performance server has both a system disk and a separate data disk(s). You can retain the operating system and configuration information from your system disk by using our Cloud Servers imaging feature, but it will not save any information from the data disk(s). To save your data disk(s) you have the option of using Rackspace Cloud Backup, which is a granular file level backup system that you can configure to save only the files and folders you wish to keep. Another option would be to run your Performance server without using the included data disk(s) at all. All data necessary for running your applications could be saved on a Cloud Block Storage volume which could be mounted to your Performance server.
Using Cloud Block Storage instead of a Data Disk
Cloud Block Storage presents data volumes to your server. Volumes are the detachable block storage devices that expand the storage capacity of your server. Think of a Cloud Block Storage volume as you would an external USB hard drive. Like an external USB hard drive, a volume can only be attached to one server at a time, but it retains your stored data even when it is not attached to a server. One limitation of using Cloud Block Storage is that you can only attach a volume to a server in the same geographic region as your volume exists.
How Do I Backup the Data Disk?
Since system images will only save data from the system disk, you will need to use the Rackspace Cloud Backup service to retain information from the data disk(s). Cloud Backup is a file level storage solution, and the increased throughput of Performance servers allows more efficient access to backup storage.
What About Scaling and Resizing?
One of the biggest differences between the current standard offering servers and Performance servers is that Performance servers cannot dynamically resize. With the rest of the OpenStack community, Rackspace is reducing support for this feature because it does not align with the industry-standard method of scaling. Rather than resize one server for vertical scaling, we instead recommend employing horizontal scaling, or adding/removing the number of servers managed by a Cloud Load Balancer to provide adequate resources to suit your needs.
If what you really need is a vertical scaling solution, either Rackspace Cloud Backup or Cloud Block Storage can be used to provide you with more data storage options. An example of how this might work would be if you saved your application data to a Cloud Block Storage volume and created an image of your system disk to save your OS configuration. Then you would build that image to a larger Performance server instance. Once active, you would then attach the Cloud Block Storage volume with your application data to the new instance.
If using Rackspace Cloud Backup, you would install the Cloud Backup agent on your current server, backup the necessary data, and then provision a new server from your system disk's image. Once the new server was active, you would connect to Cloud Backup to retrieve the saved data.
© 2011-2013 Rackspace US, Inc.
Except where otherwise noted, content on this site is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License