Cloud back-up and disaster recovery: You do need both

As more organisations become comfortable with cloud-based back-up solutions, it is becoming increasingly important to understand exactly what level of recovery their solution provides.

July 30, 2013

By Gareth Tudor, CEO of Altonet

As more organisations become comfortable with cloud-based back-up solutions, it is becoming increasingly important to understand exactly what level of recovery their solution provides. While backup is an important part of Disaster Recovery (DR), it is only one component of a full DR system. Just because you have a backup solution in place, does not necessarily mean that you have DR. For a DR plan to be successful, a number of other parameters and components need to be taken into consideration.

The goal of every DR plan is to get the organisation back up and running quickly in the event of a disaster. What, how and how often backups are scheduled is an important part of this equation: they will help determine how easily and comprehensively you can recover your operational systems in the event of a disaster and will also impact the rate of recovery, which can directly correlate to potential business loss.

The benefits of cloud-based backup
Cloud-based backup is appealing because it minimises the effort associated with traditional backup processes. Traditional backup typically involves the use of hard disks and/or tape storage, and the subsequent removal and storage of this media offsite. With a cloud-based backup solution, an onsite appliance performs an internal backup and then uses a Wide Area Network (WAN) connection to store the data offsite “in the cloud”. Once the first set of internal data is stored, only incremental backups are necessary, minimising the cost per gigabyte of sending that data to offsite storage. Backed up data may be replicated on a number of cloud servers, depending on the risk profile of the organisation.

There are a number of advantages to cloud-based backup. The backups are automated, so complexity and human error are eliminated as staff need not get involved. A central Web-based console provides status notification, management, and the ability to configure and schedule backups from anywhere. Data is propagated via the WAN, so is safely stored offsite without needing to add risk to the process through physical handling of storage media. There is also the elimination of effort and risk in terms of scheduled media delivery to or from physical offsite storage.

To maximise these benefits however, organisations need to understand how a disaster may impact their operations – and the role of backups. Data backups may allow the organisation to restore lost data, but this is not enough when disaster of any type occurs, be it a fire or other natural disaster, hardware failure, a virus, human error or a security breach. When this kind of disaster occurs, system state files are needed to restore the servers back to “current state”.

Prioritising, assessing, scheduling
In the event of a disaster, for the majority of businesses the priority is to keep revenues flowing and customers happy. To do this, key systems need to be made operational as soon as possible: communication and customer engagement channels need to be up, transactional systems need to be accessible, and customer databases and logistics and tracking systems need to be functional.

Determining this is the easy part. The not-so-easy part is knowing what kind of “hit” in terms of data loss the business can take on which systems, and putting in place a data backup and recovery solution to meet those needs.

Three key objectives can help organisations decide what will work for them:

  • Recovery Time Objective (RTO) defines how fast you need to find, access and retrieve information – that is, how fast does the organisation need to be up and running? A key challenge related to cloud-based backup and recovery is that recovery is dependent on the speed of the network. The capacity of last mile connectivity thus needs to be considered.
  • Recovery Point Objective (RPO) defines the point-in-time (PIT) that needs to be recovered and is a measure of tolerance for data loss.
  • Recovery Granularity Objective (RGO) defines how granular recovery needs to be: at a file level or at a transaction level (continuous data protection). A backup update can be made every 15 minutes on key systems or every eight hours on databases.

Cloud-based backup and recovery providers need to leverage best practice data recovery concepts like these to assess and deliver on the needs of clients who regularly backup and archive – from a few gigabytes to many terabytes of data – using different strategies. It is important to mitigate risk around data loss, not only for the business but also for clients and other stakeholders. As regulations around data usage, management and storage become stricter, the principles and practices of disaster recovery will become increasingly important to understand. Backup is a core component of disaster recovery, and as such, is a good place to start.