by Nick Mueller
Disaster recovery (DR) has been a concern for IT teams as long as businesses have relied on computers. The issue now in 2012 is that many of the techniques for delivering DR, such as tape backups, hot sites and bare metal restores have their roots in technologies that are rapidly becoming irrelevant. Virtualization, ubiquitous broadband connections and a more distributed data footprint have given rise to new requirements and recovery options.
Online or “cloud” backup solutions that provide continuous data protection are the IT industry’s response to these developments. The available solutions vary widely. Individuals and very small businesses can use low-cost consumer grade services to automatically back up their files and share them between connected devices, for as little as $5 a month. Large enterprises tend to build their own data centers or use colocation facilities for continuous replication, failover and disaster recovery over dedicated connections. This is usually referred to as a “private cloud.”
Mid-sized firms, those with approximately 20 to 2,000 employees and usually fewer than 10-person IT teams, require a different approach. Since they lack the resources of large enterprises, they need something that combines the ease of consumer online backup services, but with enterprise-grade features. For example: the ability to back up servers, storage arrays, PCs, laptops, mobile devices and all the associated data types. Requests to restore data might include anything from a single email to an entire data center, so being able to centrally manage the entire backup and restore process is important. And those processes should be secure, auditable, and the backups themselves should be backed up.
The best way for a mid-sized firm to obtain this is to find an online backup and recovery vendor that delivers Backup as a Service (BaaS). Below are the three essential elements of the best enterprise-grade online backup solutions.
Not all recoveries are the same, so the backup system needs to accommodate all possible scenarios. The most common recovery request is for a single file that has been lost, accidentally deleted or corrupted. The next most common recovery request is for the recovery of an individual file directory or server. Virtualization, RAID and local data replication systems reduce the need to perform complete restores, but they are still sometimes necessary. The least frequent, but most catastrophic, recovery involves an entire data center site. This scenario is mostly often caused by a natural disaster.
A recovery solution must be able to address all of these scenarios with an understanding of the likelihood or frequency of occurrence as well as the business impact. The recovery time objective (RTO) and recovery point objective (RPO) will vary by an organization’s needs and there is a tradeoff between cost and higher recoverability. This means it is important to set the appropriate RTO – how long it takes to execute a recovery - and RPO – how current the data restored must be - for each type of device, data and location.
Shorter RPOs – for example, continuous replication, as opposed to nightly backups – demand greater server and network resource utilization during peak usage periods, so it’s not always practical to minimize the RPO and RTO for every data source. A customer order transactional database typically requires a much shorter RTO and RPO than an individual employee’s workstation. Industry surveys have shown that while approximately 70 percent of mid-size organizations have an RTO of 12 hours or less, only 35 percent of those companies met their RTO consistently.
No On-Premise Appliances
The level of data protection offered by the best enterprise-grade online backup vendors is so good, there’s no logical reason to maintain a disk backup appliance in an often crowded SMB server room or data center. Using a backup appliance as a second disk array for on-site backups while, at the same time, functioning as the organization’s gateway to the cloud creates a single point of failure with all the associated risks.
In a failure scenario, the failure of the appliance also results in the inability to access even a single file of backup data in the cloud. This is due to the proprietary nature of how appliances and their bundled software handle backups and transmit data to the cloud. High levels of built-in redundancy are prevalent in the IT industry precisely to avoid scenarios such as this, but with on-premise cloud gateway solutions, fail-safe redundancy comes at twice the price since the only way to protect against a failed appliance is to own two appliances.
One of the advantages of backing up data online in the cloud is not having to buy, install, manage, and update hardware. Enterprise backup and data recovery solutions that are delivered as a service eliminate the single point of failure issue and provide access to backup data through a web browser. Your data, and the recoverability of the data, is protected — even in cases where the primary data center is without power or has been destroyed by a disaster such as a tornado.
A critical aspect to choosing an online backup and disaster recovery solution is how to integrate it with an existing on-site backup system. The most practical approach is to use a solution that has little or no client-side footprint or requirement such as heavy agents, local appliances or recovery servers; this makes things easy since the online backup not only doesn’t have to integrate but also becomes another layer of protection. Once they’ve gotten used to the powerful features true enterprise-grade online backup provides, some businesses may decide they don’t even need a local backup copy.
Here are the specific benefits a winning solution should provide:
- Multiple Platform Support. Files should restore to any platform,
not just a duplicate of the server that created it; this provides flexibility
in upgrading systems or using alternate equipment at a DR location.
- Ease of Set Up. The backup agents, once installed, should be able
to be configured with just a few clicks. The only configuration required
involves determining what data is being backed up, and when should the backup
- Browser Access. Both administrators and users should be able to
access the backup through a browser rather than a specialized interface.
- Ease of Restoration. Single files and folders should be able to be
restored by just dragging them from backup, through a web browser, onto the
workstation or server. Restoring servers should be as simple as mounting the
drive from backup.
- Self-Restoration. End users should be able to restore their own
files from the remote location, through a web browser, without having to
- Lightweight Agents. To minimize disk and CPU load, as well as
network traffic, any software agents should be lightweight. The agent’s main
purpose is to detect any changed files or blocks of data and send them to
backup at the appointed time.
- Automatic Validation. The DR service should include immediate verification of all backups when they occur, copying those backups to multiple servers or locations for added reliability, and regular byte-by-byte checks on the backed up data to ensure it will restore properly.
These are the elements and features to look for in an online backup and disaster recovery service. Of all the companies offering solutions in the DR space, no two are identical; the key is to make sure that the solution you chose will meet the particular business requirements of your organization. Done right, Backup-as-a-Service is a secure, scalable way to protect critical enterprise data and ensure business continuity under any circumstances.
Nick Mueller, who has been writing and telling stories about technology via blogs, social media and content marketing since the days when the BBS reigned, is a corporate reporter for Zetta (Sunnyvale, CA). www.zetta.net