When did it become acceptable to throw I.T. best practices – such as backing up your data – out the window? When it comes to SaaS (“cloud”) applications, there seem to be two types of people:
- Those that think that archiving is backup
- Those that don’t think they need to either archive or backup their data
A quick review of the differences between backing up and archiving data reveals where some of the confusion lies.
Backup. It’s all about the recovery. Backups protect your active data (files, etc.), allowing you to recover individual items (or, inc the case of Google Apps, a full user account) to a point in time before the data corruption occurred. Backups are all about RPOs (recovery point objectives) – the point in time to which you want to recover, and RTOs (recovery time objectives) or how long should it take to recover to your desired stable state. The more recent the RPO required, the more expensive the solution. The same applies for your RTOs.
In addition to RPO and RTO requirements, many companies also demand that backups be stored off-site, in a second location.
If you had an email server (Microsoft Exchange, etc.) on premise, you would back it up, wouldn’t you? There is no reason why you shouldn’t continue to backup your data once you move to a SaaS application. Many of the same risks of data loss still exist in the cloud. If you still need to recover lost of corrupted data quickly and reliably, you need independent backup.
Archiving is all about legal discovery. Archiving focuses on preserving inactive information, often in response to legal requirements or company policies. Some companies have to archive their data; others choose to. Archiving is all about preserving data in case you need it in response to some unanticipated legal or regulatory event.
A secondary benefit of archiving is that it can allow you to move data from more expensive (“near-line”) storage to less-expensive storage systems.
Why do both? Isn’t archiving data enough? There are several reasons why relying on an archive only is dangerous. Firstly, archives don’t protect all of the copies of the information. An archive is designed to preserve a single, index record. And secondly, good archiving solutions are optimized for the discovery process. They aren’t designed to allow you to easily restore individual items back into your native applications.
The business requirements to either backup or archive your data do not change when you move to the cloud and adopt SaaS applications. Would you buy a hybrid car without airbags?
It’s all about using the right tool for the job. If you want to be able to recover from user errors and go back to a certain point in time (to meet your RPO requirements), then you should be backing up your data. Regardless of whether it is located on premises or in the cloud. Archiving solutions are not set up to meet these requirements. By the same token, backup solutions are not designed to preserve audit trails, so they should not be relied upon to meet the most stringent compliance requirements.
Backups are about recovery. Archives are about discovery. You may need one. You probably need both. Moving to the cloud doesn’t change that.