Archive functions are changing and traditional tape backup software may not be the best method of loading data into an archive.

A few days ago I talked with Yosemite Technology CEO George Symons about backup and archiving and metadata needs. Archives based on traditional tape backup models are going to need more metadata in order to meet compliance and legal discovery needs.

This prompted Bridgehead Software MD Tony Cotterill to send me this mail:

"Imagine SRM (storage resource management), archive and backup working together in harmony. SRM runs every night, deleting what is not required and archiving what is, leaving behind only a copy of those active files that are needed by users and applications.

"The archive process ensures that only a single instance of the file ever makes the archive. It also directs different file groups to different media sets depending on their need to be accessed and their security requirement. Also as these files are archived, they are fully indexed both by their attribute data and by their content.

"The result is a completely indexed (attribute and content) store of files whose whereabouts is fully known. This information store is fully searchable by the user (only his own files) and the organisation (the entire file system). Add compression and encryption to the process and you have a very compact and secure file store.

"It only remains to take care of the files that remain on primary store. It is worth noting that this is only 20-30 percent of the original amount of data. This is backed up by the integrated backup tool - not for the purpose of file recovery (these are in the archive) but purely as a point-in-time copy of the primary store for disaster recovery. This backup is quick because it can be done at the block level ignoring the file structure.

"This is not stargazing. We are doing this now with customers all over the world across the public and private sector and have been for a while now. Have a look at this interview with Dave Lipsey at Ordnance Survey on our site and there's a white paper on the Protected Data Lifecycle Management (PDLM) approach, which we have pioneered over the years. This provides a pragmatic approach to the challenges of ILM (information lifecycle management)."

Worth a look? I think so.