Disaster Recovery — 23 August 2010

Marc V. Mombourquette, product marketing manager at Acronis (www.acronis.com), says:

How do you know your disaster recovery plan is up to date?

First identify and isolate a Recovery Time Objective for each application. For instance, if you suffer a flood and your Exchange or SQL server applications must be accessible by both internal users and partners on a 24/7 basis, your recovery time objective (RTO) for returning these functions to full power will probably be measured in minutes rather than hours. If that’s the case, you’ll want to work from disk image-based backups, with their very fast recovery performance, and you’ll want to establish either an offsite for online recovery services facility to get you through the crisis while you rebuild from the flood. A human resources application, on the other hand, in which the data changes far less frequently, can probably be assigned a less urgent, less costly RTO without substantially impeding daily operations.

When faced with circumstances like regulations and data growth – and don’t forget the relentless push to lower IT costs – traditional file-based backup systems won’t be able to keep up from either affordability or performance perspectives. They will require more personnel and monetary resources than most small and medium-sized businesses can afford. To meet these challenges without overwhelming your staff or budget, IT departments should consider a backup and recovery solution that:

  • employs image-based backups that allow granular recoveries of files and folders as well as bare-metal recoveries.
  • includes a scalable architecture that seamlessly manages change without disruptive ‘forklift’ upgrades.
  • features a user-friendly interface that drives down training requirements and increases productivity.
Handling performance requirementsIT organizations must be able to recover from all sorts of failures, whether it’s a lost file or folder or an entire machine, in much shorter timeframes than were set even a few years ago. Today’s recovery goals are measured in seconds or minutes, rather than the hours or days. This effectively rules out tape backups, with their inherently slow recovery times and the added frustration of  occasional tape drive failures. And it rules out file-based backups, which leave IT organizations with no way to quickly recover failed workstations or servers. To address performance requirements, IT administrators should consider a combination of image-based backup solutions and system virtualization if they’re not already moving in those directions.

Image-based systems  take a snapshot of a machine while it’s still running, eliminating the need to take it offline and multiplying the speed with which you can complete a restore. Because everything – data, applications and all settings –  is  backed up in a disk image, you have all the ingredients needed to restore anything from an individual file to the entire machine itself, in just seconds, to either a physical or virtual machine. It’s easier to enhance the performance of physical and virtual machines when your backup and recovery software manages both through a common user interface. Many don’t, forcing administrators to set up and administer parallel backup and recovery schemes, taking them away from more proactive work.

Handling capacity requirements. Because more and more applications are deemed mission-critical to the work of both internal and partner personnel, more and more volumes of data are considere critical to an organization’s success. As data storage needs increase, IT administrators have real concerns about where to put it all. That’s why deduplication is one of the most effective ways to address ballooning data storage requirements. Until recently, the cost of proprietary hardware-based deduplication, the sheer size of its hardware and its major installation requirements took it out of consideration for small and mid-size organizations. The relatively few3f organizations who acquired usually only deduplicated server data, despite the fact that workstation data frequently represents half of the entire data owned by an organization. This all changed with the advent of software-based deduplication. It substantially lowers the threshold for purchase, using inexpensive commodity hardware rather than expensive, proprietary appliances, so IT organizations can afford to deduplicate server and workstation data. Look for a solution that is image-based (faster), can be fully integrated with your existing backup and recovery solution (easier to administer), and that supports full, incremental and differential backups (more flexible recoveries).

Changing requirements. When the your IT organization’s requirements begin to shift,  and you find that your backup and recovery solution can’t keep up without significant, time-consuming IT intervention, you may be working with the wrong  solution. The best backup and recovery solutions on the market are highly scalable to meet expanding data needs, allow you to set policies to automate backups for whole groups of machines at once, and feature the flexibility you’ll need to complete transitions smoothly.

For instance, if you’re beginning or accelerating your investment in virtual systems, look for a data protection product that doesn’t limit you to a specific virtual machine platform. Cross-platform moves are becoming common among the four major VM platforms on the market (VMware ESX and Sphere, Microsoft Hyper-V, Citrix Xenserver and Parallels), and you’ll want to be able transition seamlessly between physical and virtual platforms as your computing needs change. Look also for agent-less backup, which reduces administration requirements substantially. 

IT administrators are constantly faced with change, so it makes sense to move to disaster recovery products that adapt fluidly as change presents itself. Such products should allow you to work smarter, reduce IT administration costs and do everything required to protect your data. Acronis foresees a convergence of capabilities that will allow small and medium-size businesses to enjoy all the performance and functional richness found on the enterprise level, but at prices any organization can afford.

Inexpensive, yet powerful deduplication, image-based backup and instant recoveries, and enterprise-class management  capabilities will be enhanced by low-cost remote management of resources and data security offerings. Further, we see the advent of cloud-based online services that allow immediate recoveries to remote locations in the case of a disaster. Combined, they will blur and then erase any functional and performance differences that exist between the disaster recovery capabilities enjoyed by large IT shop and those available to smaller IT groups. Any differences that exist now will have disappeared within the next two years, heralding a new era in disaster recovery for small and medium IT organizations.  

Share

About Author

(0) Readers Comments

Comments are closed.

Visit Us On FacebookVisit Us On TwitterVisit Us On Linkedin