The COW ate my data

How many times have you saved a document only to realize later that you really need that old version back?  Saved on top of an existing but different document? Deleted a file by mistake?

Of course, most of these scenarios are covered by various backup methods (you are performing regular backups, right?) but suppose that something happens between backup cycles or that regulatory requirements specify retention of all versions of a document, etc…  Enter a ‘versioning file system’.

A versioning file system automatically creates snapshots of files each time they are changed/saved.  What you see in your directory listing is the latest version but behind the scenes are maintained historical versions that track all changes made to the file over time.  For example, you can easily recreate the file as it existed last Tuesday at 2PM, a month ago, or 2 years ago.  All this without tapes!

None of this magic is free; the cost is increased disk space utilization.  However, only the changes to the data are stored, not entire copies.  The amount of overhead is dependent on the frequency and amount of changes to a file.  A solution like this is not appropriate for a highly-dynamic file such as a database or email inbox but is a solid solution for document files (MS Word, Excel, Powerpoint, etc.).

Welcome

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.