IT consultants require a solid backup strategy, because sometimes we’re responsible for backing up dozens or even hundreds of client servers. The more systems there are, the higher the likelihood of failure. When servers fail, it’s critical that you have healthy image-based backups from which to quickly recover clients.
Image-based backups provide you with these vital safeguards:
- Hardware independent recovery
- System state data
- Windows registry information
- “Forgotten” files
- Bootable backups
- Faster recovery
Let’s take a look at each of those elements a little more closely.
Hardware independent recovery
When a server unexpectedly fails, the ability to recover the server’s image to another machine quickly proves the difference between prolonged downtime and an inconvenient outage. Recovering a server image from one brand and model server to another has enabled my consultancy to keep more than one client afloat and operating following a catastrophe. We’ve even recovered servers to high-powered desktops, if only temporarily, to buy time until a permanent server replacement could be ordered, shipped, and installed.
System state data
Many traditional backups back up only the information a client creates. That’s well and good, because that’s critical data. But if you must recover a client’s server from total failure or corruption, healthy system state data is required to rebuild the system.
Information is also needed about user accounts, folder structures, permissions, applications, and other critical settings. Many consultants inherit server backups in which this information is omitted or cannot be backed up using the current backup solution. With properly operating image-based backups, you can rest assured that important system information isn’t being neglected and is safely being collected.
Windows registry information
Many backup operations cannot or are never set to back up Windows registry information necessary to properly boot a server and load the operating system with no errors. What’s worse is many backup programs (even those that collect registry data) might miss critical system files or drivers.
My consulting office occasionally encounters client servers whose RAID arrays corrupt or that experience OS corruption preventing Windows from successfully booting following electrical outages, update installations, and similar common occurrences. Image-based backups frequently enable my firm to quickly recover these clients’ systems, as image archives provide ready access to registry data, system files, and even drivers.
Military officials talk about unknown unknowns, or the things you don’t know you don’t know that are going to come back to haunt you.
Clients who install new applications that store data in unique locations and clients who fail to alert you to newly created data directories run the risk of creating a crisis if data have to be recovered from a previously scheduled backup that hasn’t yet been updated to collect the new information. Image backups by design almost always back up any data stored on preconfigured volumes; this means that forgotten file surprises are virtually eliminated.
Unlike traditional backups, image backups typically include utilities that make it possible to create bootable images. Why is that important? When RAID arrays go south, motherboard controllers fail, hard disks implode, drive updates wreak havoc, and other errors occur, being able to boot a previous version of the OS can prove critical in recovering a server or simply accessing key files, drivers, settings, and configuration information.
Image backups’ greatest benefit is speed. In the past, when servers failed, technicians would obtain a new chassis, reinstall the OS, locate and load drivers, recover system state and other data from the backup, download and install OS updates and service packs, reinstall applications (after tracking down long-lost media, license keys, and registration information), and then recover files and other information from the old backups. The process took several days.
With a hardware-independent image backup, a savvy engineer can recover an entirely failed server in less than an hour or two, assuming a chassis of some sort is available. But with an image-based backup, in an emergency a laptop can even be used to recover a small business’s server; I know because I’ve seen it done while waiting for a permanent replacement. While it’s not ideal, it has kept companies running. The time-savings alone more than justifies the cost of image-based backup software.
Don’t get caught by surprise
At least one paper suggests significant technology department time is spent troubleshooting failed backup operations (PDF). And, just because you receive a successful backup confirmation, don’t assume that means an image backup can successfully recover a failed server.
You should educate clients about the importance of paying your consultancy to monitor backup operations daily. I also recommend encouraging clients to invest in paying your office to regularly take image backups offsite to test how well the image backup can recover an actual secondary system to bootable operation.