Data management problems are growing alongside data volumes, and cloud-based
storage may be an option. Many businesses have access to plenty of bandwidth so
can move primary data into the cloud for backup and recovery purposes, paying a
monthly fee and avoiding capex.
Cloud storage, related to the concept of storage as a service, is not new but
has become something of a land grab. Just like at any new frontier, the rules of
engagement may be unclear and hype may surround any opportunities therein.
Cloud storage faces a plethora of international regulatory hurdles, around
corporate governance, compliance, and data privacy. And it used to be a
fragmented niche market, with few nationally recognised standards.
Backing up and recovering data from any source, including the cloud, demands
resilience, reliability and availability. Cloud storage relies on internet
connectivity and bandwidth so still has much to prove in this respect.
In fact it is probably bandwidth that is most likely to hinder the widespread
adoption of cloud storage. As data volumes grown, bandwidth issues become more
critical.
UK consumers who once relied on a 28.8k modem now access pipes delivering
around 5Mbps. Data growth has exploded from perhaps 100MB per user to well over
a terabyte. Off-site storage for primary data in many ways is becoming less
rather than more feasible.
Meanwhile, any company storing data in the cloud depends on its cloud
provider for the safety and availability of its data. Addressing reliability and
security concerns are paramount if companies are to trust cloud storage.
Once initial backup has been done on site, a copy can be hosted in the cloud.
Technologies that support automated incremental backups are best suited since
only the new data needs to be backed up in the cloud. This incremental approach
saves bandwidth and time.
IT managers need to familiarise themselves with their cloud provider’s
policies and methods, and understand the consequences of a worst-case scenario.
Geoff Barrall is chief executive and founder of
Data
Robotics
Do you agree?
Have your say on this article