Keep backup simple and do it well
A storage vendor lesson
Comment There is really nothing much to backing up data - or there shouldn't be. The trouble is there is so much of it to back up these days and, with 24/7 working, no longer an opportunity just to shut down until the job is finished.
Then, for a large enterprise at least, there is the possibility of gathering data across a large network which includes WAN as well as LAN connection.
So we read a lot about continuous data protection (CDP). Close examination reveals this title is a little flattering since, even if you can ensure the ability to recover from every single transaction (you cannot), this would still not, on its own, ensure that the problem that caused the outage would not repeat itself straight away.
I feel more comfortable with Double-Take Software's slightly more modest claim of continuous data replication and I concur with Double-Take's UK sales and marketing directing Ian Masters when he asks: "Why would you need to restore to any point? It slows things down and recovery is not likely to be any quicker anyway."
What most users would like to be doing is to cut through the back-up and recovery complexity, letting the software handle all that transparently, so that it "does exactly what it says on the tin" - and does it well. That may not sit well with the major storage management vendors who obviously want to sell lots of different products from their extensive software (and sometimes hardware) portfolios.
There is also the 80/20 rule; nominally 80 per cent of the benefit comes from 20 per cent of the functionality. Concentrate on getting the most important things right and you may then have time to consider what to do next and be able to see the wood from the trees as well. There is little that's more important than ensuring regular back-ups with the ability to restore reliably and efficiently if needed.
That may help explain why Double-Take Software has been doing solid business recently. It focuses on doing this back-up and restore function - and doing it well. The software is entirely Windows-based, requiring access to Windows I-O (although its support for VMWare now allows some flexibility here), with Exchange and SQL Server support. But this limitation only contributes to Double-Take being inherently more focused than most of its larger competitors whose products may for instance support Linux, UNIX flavours, and a couple of more proprietary operating environments.
Masters believes Double-Take's software scores for resilience, recoverability and ease of use - and excels in WAN environments through byte-level access, compression support and the ability to set bandwidth levels. As copying is asynchronous a line drop can mean delays for re-synchronising, which is a down-side; conversely, it has no problem handling one-to-many, many-to-one and many-to-many hardware systems backups, with failover for DAS, NAS and SAN. As Windows utilisation is fairly low - say 20 per cent - its background operation using 3& - 5 per cent of CPU usage is also insignificant.
However, I am in danger of slipping back into describing software complexities when these are hidden from the users' view.
So, returning to the theme of simplicity, the software is sold entirely through the channel and licensed on a simple "per server" basis, which means it often works out at lower cost than the competition. Set-up is wizard-driven and a dashboard can be used to visually analyse server throughput or watch data bandwidth usage in real time.
In other words, this is approximately what the average business user should be seeking in backup and restore software. It does what it says it does while hiding the underlying complexity.
What has been the pay-off for Double-Take? It now has well over 10,000 user companies (over 30 per cent of these in Europe) including some big-named enterprises. So evidently, despite some major vendors' bells and whistles of functionality, these enterprises obviously decided a separate application for this straightforward task was a better choice for them.
Sadly, I doubt the big boys will learn this lesson from Double-Take. Past experience tells me it's more likely they would ponder the merits of snapping up the company, its software - and lots of users. Then what would happen to that focus?
Copyright © 2007, IT-Analysis.com
Sponsored: RAID: End of an era?