Tuesday, October 20, 2009

The real issue is replacing the data that’s lost (or corrupted, etc.) and restarting the affected applications. The goal should be Recovery, with backup being just a part of the process to achieve that goal. When IT people talk about what worries them, it’s usually whether their data protection system is working and whether it will be able to get them through a future data loss event.

There’s an old saying in the real estate industry when referring to the value of a house.  “Location, location, location” means where a house is located is the single most important determinant of its value. Using this analogy, IT is really saying the thing that keeps them awake at night is “recovery, recovery, recovery”.

For many companies, they focus on the process of backup instead of the goal of recovering data. It’s understandable, without reliable backup, recovery can’t happen.  However this focus leads to outsourcing the backup process to a backup service provider or choosing to do it in-house. But, again, since backup is just a part the data protection they’re looking for, maybe what’s needed is a Recovery Service Provider.

Recovery Service Provider

A recovery service provider, like Simply Continuous, should be better qualified to complete the solution that backup starts. A recovery service provider also provides systems to deal with some of the other areas that could go wrong when you try and replace lost data with a copy from backup and restart the failed applications. These can be issues with the integrity of the storage infrastructure or data center used or technical support issues that arise when you experience the loss event. They also need to provide simple and efficient access to data and flexibility in the way restores are accomplished. Finally, a recovery service provider should offer a Service Level Agreement (SLA) for recovery, guaranteeing the timeframe required to restore data.

Applications

In addition to data recovery, applications also need to be restored, promptly, to facilitate continuance of business. After all, data is only as useful as the application it’s associated with. One of the ways a recovery service provider can do this is to host key applications in the same secure data center that stores the data, with virtual machine images of critical applications loaded onto hosted servers. In the event of an outage at the primary data center these applications can be started quickly. As further assurance of minimal downtime, they can also provide the ability to perform on-demand recovery testing to provide proof to you that you can recover when you actually need to.

An alternative to this service is to provision your own remote-site data center with dedicated hardware and have IT resources available on-site to get these applications running. Or, these backed up data sets would need to be transported to where the infrastructure is located and the applications restarted from there. These may not be feasible options, if the cost it too high, or if relocating storage hardware takes too long.

Backup Service Providers

When a company’s decided to upgrade from ‘backup’ to ‘recovery’ there are a number of options available. However, most are really classified as backup service providers since they generally don’t focus on recovery or provide application support.

Turnkey backup and DR services are perhaps the original outsourced data protection solution. But they usually entail extensive consulting, infrastructure, bandwidth and implementation engagements - as well as large, on-going monitoring and support fees. For some organizations, this can be too costly.

Some companies offer storage capacity ‘in the cloud’, meaning they can store primary backup data directly or take a copy of data already backed up and hold it off-site. They typically use change-based algorithms (like deduplication) before transferring data to reduce the bandwidth required. These systems can get the backup part done, but can run into problems when large restores are needed.

Other solutions involve a ‘hybrid cloud’ model, where they capture backups on-site, in a disk-based appliance and then transfer to the cloud. Although viable, many require that the organization be at the point where they are ready to replace their backup application something many storage managers may not be willing to do.

Another option is to keep the backup infrastructure in-house and look for an off-site location to send data for DR. A colocation facility may be used, if the company doesn’t have a suitable remote site. This choice can leverage existing backup infrastructure but also involves significant upfront and on-going effort on the part of IT, as well as expertise - not to mention the absence of a “recovery” SLA.

Overall, the emphasis must be on replacing lost data and restarting IT processes, not just backing up data and getting it off-site. Compared with a recovery service provider, these solutions may still fall short, especially if monitoring, on-demand testing or application support is needed.

Paying for a backup service provider when they only provide part of the solution that’s needed, is like using a landscape service that doesn’t haul away the grass clippings or taking your car in for regular maintenance and being told they don’t check brakes.  Hiring part time help is one thing, but using a traditional backup service provider can be like buying part-time insurance. When needed, it may not be there for you. With a recovery service provider, the focus is on recovering data, restoring applications and providing the guarantee that you’ll be back up and running - not just backing up your data.

Eric Slack, Senior Analyst