Don’t write off disk

Virtualisation is quickly changing the way businesses operate their IT departments, with many companies already jumping on the virtual bandwagon. The high adoption rate only emphasises the ability to combine capabilities and streamline processes. However the benefits of physical appliances have been widely known for some time. As new technologies are implemented it’s important to look closely at the advantages of traditional storage such as disk. And contrary to what you may have heard, disk is anything but dead.

This isn’t the view of a company wedded to tape. With the use of object storage disk for retaining all unstructured archived data, we’ve already seen a new lease of life whereby corporations are able to store and access everything in ways which were previously not possible. And by moving all unstructured archived content to a disk-based object storage platform, you accomplish several things at once. First off, you immediately reduce the burden on your primary storage resources, as well as improving the performance of your backup operation. At the same time, you’re able to free up network bandwidth that is involved in your backup processes.

What strengthens the case to opt for disk now more than ever before is down to two key ingredients. First, today’s next-generation object storage technology includes some level of forward-error-correction technology that enables objects to be spread across components to accomplish durability much like what RAID technology does for drives within an enclosure. However, some implementations of object storage can go way beyond spreading data across drives in an enclosure. Data can now be spread across enclosures and even across geographies, and in doing so the user can ensure extremely durable archive data protection. Protection against component failure, against enclosure failure, and even against site disaster.

Previously, when users thought about doing this, they’d think about moving archive data from primary storage to less expensive disk or maybe even tape. But they’d still insist on at least two copies — and in many cases three -– to ensure their archive data didn’t disappear in the event of failure or disaster. Now a single instance of data on object storage technology provides better durability than RAID and disaster resilience as good as tape, so users have the best of both worlds with a single instance on a single technology -– object storage.

The second ingredient is common management tools. There’s been a lot of discussion about the convergence of backup and archive, and the message is starting to gel with many that convergence is needed at the management layer.

It’s not a convergence where backup data and archive data are the same, and it’s not the case that backup storage and archive storage are the same. It’s really about ensuring the backup management layer and the archive management layer are completely in sync, or when possible, are actually the same management layer too. With that, policies can automate the treatment and movement of backup data differently from archive data and the system can then place archive data where it is optimised for its use and backup data to where it is optimised for its use.

So further investment in disk can now continue to hold the vast amount of archive data corporations want to retain and have access to long-term. With next-generation object storage technology and comprehensive management tools now available, users will no longer have to add the additional costs associated with second and third copies to back up the first instance. This will allow users to expand their disk storage much more cost effectively since a single instance of data will suffice.
With that said, disk isn’t quite ready to written off just yet.