Storage Tip: Protecting PCs against logical data protection problems
Send your Storage question to David Hill today! | See other Storage tips
What seems to be the problem? Users view their personal computer -- whether a desktop or a laptop or both -- as being essential to their everyday productivity. Backing up over the Internet (for example, Berkeley Data Systems or Carbonite as discussed in other storage tips) or to another drive can protect against physical problems, such as a hard disk failure. But what about logical data protection, such as against accidental deletion of a key part of a document, or the data corruption of a file for any reason. The purpose of backup is to save data so that it can be restored. That data is the version of the data that the backup was taken (which may or may not be the latest version). No guarantee that that version will suffer from a logical data problem.
And logical problems are the type of operational problems that can ruin a user's and your day. For example, accidentally deleting key paragraphs from a complex proposal and then saving the document would leave a version that backup may or may not be able to restore depending upon timing. Your challenge is to help your user community solve their logical data protection problem.
What do you need to know? Snapshot technology is one approach to solving the problem. A snapshot creates a virtual copy of a disk volume at a designated point in time. At the time the snapshot is taken the original disk image and the snapshot are identical. Thus no additional physical space is required at that instant. The original data and snapshot diverge as writes change the original data.
Since taking snapshots is really not intrusive performance-wise and is typically unlikely to have a significant storage usage impact, then multiple snapshots can be taken over time. Each snapshot is a virtual copy of the data at a point in time. If a logical problem occurs, you can restore to a version of a file or system prior to when the problem occurred.
Rollback Rx, a product of Horizon DataSys, illustrates this snapshot strategy. (Horizon DataSys briefed me as an industry analyst on its Rollback Rx product.) If allowed, individual users can roll backwards and forward to any system snapshot without requiring IT support. But there is also an Enterprise edition that enables IT to remotely manage snapshots and rollbacks through a Web-based interface if that is the preferred mode of operation. And Rollback Rx is not limited to restoring individual files, but can reverse any systems crash (even if Windows cannot startup). Rollback Rx can also help with other key logical problems, including a failed software installation or a botched update.
What can you do about it? You always run the risk of losing some data from a logical data protection problem (which is your recovery point objective). However, if you use snapshots at tight enough intervals, the recovery point becomes much smaller. However, users can easily roll-back to a previous snapshot, yet still be able to access files from a future (or previous) snapshot. With this technology snapshots do not have to be in "tight" intervals. (i.e., you can roll-back to yesterday, yet still access your most up-to-date data).
The use of scheduled image snapshots is therefore one of the technologies that IT may want to look into to help protect their end users from logical data protection problems on personal computers.