A caveat to cloud disaster recovery is network bandwidth and budget -- in terms of how much data can you protect and in what timeframe. The trick is to establish different recovery time objectives (RTOs) and recovery point objectives (RPOs) based on a business impact analysis (BIA) for different levels of data. For example, on my own systems a regular disk-to-disk (D2D) backup is performed to a local network-attached storage (NAS) device, along with periodic backups to removable media that is sent offsite to a vault. In addition, important and time-sensitive data is sent via network connection to my cloud service provider, including when I'm travelling on business. The idea is that I have a full copy of data in multiple locations using bulk data protection techniques while leveraging the cloud for smaller amounts of time-sensitive data.
Bottom line: If you are an SOHO or SMB, look at how cloud data protection services can either complement or replace how you're doing backup, data protection, disaster recovery and even data archiving. But make sure that your data protection via cloud is not a single point of failure.
This was first published in August 2010