|
Managing Information November, 2005 By Stephen Lawton Original pdf (168 KB)
Firefighting Through Disk ImagingTime in the IT world is measured differently than in normal society. While it is perfectly acceptable to wait one minute for the traffic light to change, it is unacceptable to wait 15 seconds for an application to launch. Seconds feel like minutes, minutes feel like hours, and when data is unreachable, a day can be an eternity. Recovering from a server disaster only intensifies these anxieties. A server that is down generally impacts on more than just the applications it runs. A failed web server can bring online sales to a grinding halt. A faulty file and print server can turn a billing department into a vast paper wasteland while clerks are unable to print bills. And a downed SQL server can make almost any job into a nightmare, regardless of what the employee does. Traditionally, IT departments would backup their server files nightly to tape libraries. Should a server fail, the IT department would have to rebuild it from scratch. Once the operating system and applications were installed and fully configured on a new disk, data would be brought back from tape and the server could be running again. This process could take days to perform, depending on the complexity of the server, the related configuration files and the location of the tapes that had to be brought back to the facility to reload the software. This assumes, of course, that there are compatible tape drives at the same locations as all the servers. Today, however, a downed server needs to be back up and running virtually immediately. While some hardware-based stor-age technologies facilitate this process, such as RAID systems with hot-swap drives, these approaches are expensive, require considerable hardware and software expertise, and are generally part of a much larger storage management solution. And, curiously, some of these solutions don't adequately address how to restore a single server — or group of servers — back to service instantly if they fail. What the IT manager needs is an immediate way to restore a server to working condition, even if the most experienced IT staff members are not available. They need a reliable, robust and straight-forward disaster recovery plan. They need disk imaging.
The Cost of DowntimeWith an image of a server's hard disk, the IT manager can restore a system to a known, working state quickly, completely and efficiently. A disk imaging product that works at the sector level, such as Acronis True Image Enterprise Server, permits the IT manager to restore a drive that suffered a logical failure, such as a virus attack or an application that has become corrupted, simply by restoring a known, good image. In the case of a physical failure to a hard disk, the IT manager can install a new, bare-metal disk and restore the image directly to the drive — no formatting or OS installation required. Daily incremental updates, coupled with complete, weekly updates, can provide an IT manager with a known, good state that is no more than 24 hours old. And because Acronis True Image Enterprise Server can operate in the background by imaging only changed disk sectors without impacting system performance, the software can make even more frequent disk images, reducing the window between a system failure and a full disk image. This approach requires the IT manager to look at backup and disaster recovery planning a little differently than in the past. Rather than considering a given number of hours or days of downtime "acceptable" when rebuilding failed server disks, that number can be exponentially smaller and measured in minutes when restoring an image. Additionally, it adds an extra layer of security and legislative compliance to the enterprise server environment. An image-based approach to disaster security — particularly where images can be stored both locally and off-site at a network-connected server in a secure facility — allows the company to comply with stringent new laws that require maintenance of email and other records. Just how much does downtime cost? According to the Meta Group, a leading market research firm, it could run into millions of dollars per hour. For small to mid-sized businesses, downtime is even more dire and can be fatal for the company. Market research firm Gartner, Inc. estimates that two out of five companies that experience a disaster will go out of business within five years as a result of the event. According to the US Bureau of Labor Statistics, 93% of all companies that experience "significant data loss" are out of business within five years. These numbers are staggering when one considers how much businesses rely on storage systems today.
Risk IdentificationThe problem with disaster planning is that almost anything can turn into a disaster. That leaky water pipe in the wall could result in a flooded server room. The gas main in the middle of the street could erupt into an inferno if the road workers' backhoe misses its target and hits the pipe instead. The key to risk identification is determining what risks are likely to happen, which will have the greatest impact on your company, and how to prevent them. Let's look at a couple of examples. Companies in Los Angeles, California, probably don't have to worry about snow damage to the roof or electrical storms causing a power outage, but they do need to be concerned with where their off-site backups are stored. Setting up a secure virtual private network to a location on firmer terra firma might make a lot of sense in earthquake country. Likewise, a company based along the Gulf Coast of Florida might not want to have its server room in the basement of the building, considering the potential for floods during hurricane season. Once risks are identified and prioritised, the company's IT department, in conjunction with senior management, needs to determine ways to mitigate the most likely problems while ensuring that these plans don't put the company at risk for a less common, but potentially severe, hazard.
Disaster Recovery PlanA well-devised disaster recovery plan not only secures corporate servers from outside influences — for example, anti-spyware and anti-virus products, as well as firewalls and encryption — but it also plans for possible internal emergencies. These might include a fire or flood, defences against an internal security breach, or simply a disk drive or server that fails due to hardware or software problems. Disasters need not always be acts of god; sometimes they're acts of software or simply negligence. Disaster recovery plans differ based on the size of an enterprise. The small or mid-sized business with 30 employees and four servers (email, web, file-and-print, and the applications server) will have significantly different needs than an enterprise with hundreds or thousands of servers spread around the world. And it's not enough just to have a plan written down in a binder no one reads; you have to practise the plan, making sure that everyone knows what to do and how to do it. This is equally true for a business of any size. Plans alone are nice, but if no one knows how to recover a failed server — and if they've never practised how to recover a downed server — then the enterprise is not fully prepared.
ConclusionServer disk drives are still mechanical devices. It's not a question of if a server drive will fail; it's a question of when. Being prepared with a disaster recovery plan in place — a plan that includes disk imaging of servers on a daily basis — is crucial to keeping these servers running with the lowest possible downtime. A disk imaging solution for servers — be they racks of enterprise servers or a single or small group of servers in a small or mid-sized business — is an insurance policy no IT manager or business owner should be without. Stephen Lawton is director of marketing at ACRONIS, INC.
| 
Acronis Inc. Doftware de Backup.
Adquirir Actualização