Cloud Backup and Restore

Cloud Backup and Restore: the Advantages

Backup and restore still remains a black art for many. Although companies accept viscerally that data is now so central to their business that it should be protected, many small and medium businesses (SMB) and mid-market organizations still take an ostrich approach to data protection. In many cases, there is no formal back up process or it is at best ad hoc; even where processes are in place systems are never checked and in reality recovery may be tricky; in others, it is felt that using e.g. email as a replication store for data is enough of a step to allow for recovering after any data loss.

However, now on-demand (cloud-based) services are coming to the fore as a consideration for data protection. As an external environment for storage such services hold much promise. For example, being off-site, they provide good protection against disasters as well as just component failure (e.g. a flood taking out a complete building as opposed to the failure of a disk drive). Such service are also “elastic” – if you need more storage, then the cloud provider can allocate it, generally on the fly, so the business is not faced with large capital bills for storage units, tape systems or whatever. Furthering the move away from capital costs, the amount of storage used is generally dealt with on a subscription basis, so making the costs relatively predictable and also controllable to a degree.

What to Consider

So, cloud-based backup and restore is perfect for the SMB and mid-market, then? Well, yes and no. There are a few issues that anyone looking to the cloud should bear in mind and ensure that they are happy with the way the provider deals with them before proceeding.

  • Minimize the amount of data. Much of the data that a business deals with is repetitive and contains a lot of redundancy. Ensuring that data deduplication is in place can reduce the amount of data needing to be backed up by up to 80%. This then lowers the costs of cloud storage – and also makes any need for restore faster after any problems. The customer can either carry this out as part of its overall data policy, or can use a provider that builds in target-based deduplication in its services.
  • Initializing the first backup. If all you want to do is backup a small amount of data, then an ADSL line will probably be OK. But, for an organization with a few terabytes of data, such an initial backup could take weeks – and hammer your available bandwidth. Better to backup to a physical storage device at LAN speeds and ship the device to the cloud provider, so that they can transfer the data to their systems at LAN speeds too. Conversely, when it comes to the need for recovery, the same approach should be available – the cloud provider copies the required images/data over to a storage device, ships it to the customer who can then carry out the full restore at LAN speeds, rather than over the WAN.
  • Continuing the backup. Once the first full backup has been created, the amount of data being transmitted over the WAN needs to be controlled. The best way to do this is for the provider to carry out “incremental” backups, which just look at what has changed over a defined timescale and create a small additional file that is transferred and stored. On a regular basis, it is best to then consolidate the original image and incrementals to a single file to make restores less of a hassle.