Information is vital in your everyday work life.
It drives business and communication. It lurks behind negotiation, powers
marketing strategy, colors decisions, comprises the core organizational records,
and calls people to action. A desire for more information is why you've come to
this Web site.
With information stored virtually as data at a
premium, not only for business uses but for compliance and legal reasons, every
company must take steps to protect data, ensuring its availability no matter
what may come. That's where backup comes in.
Backup Basics
Creating a backup of your files involves taking
a snapshot of data as it exists at that particular moment in order to create a
restore point. That copy of your data can then be accessed and reloaded in the
event of a crisis, or simply preserved for archival purposes. Backups are stored
on a different piece of hardware or even in a different physical location from
the data they are designed to duplicate.
Given the value of information today, it's no
surprise that data backup has blossomed into a multibillion-dollar industry.
(Specific products and players in the industry are discussed in the What's
Available section that follows.)
Proper preservation of data requires enacting a
backup strategy, implementing a system, and quickly restoring data after a
problem arises. Careful planning and proper foresight are vital.
Best practices have traditionally hailed tapes
as the primary option for data backup. Tape systems--costing about $2/GB--offer
the highest bang for the storage-space buck and can be easily removed and stored
offsite. Most backup solutions at a base level involve multiple versions of
tapes rotated through safe storage at multiple sites. Drawbacks to tape backup
include slow restore times and occasional reliability concerns.
In recent years, disk-based systems have become
more affordable and are increasingly becoming part of a multilayered backup
strategy. Virtual tape libraries--averaging about $16/GB--use a disk-based
process that mimics the behavior of tape systems. Like tapes, VTLs are easy to
integrate and flexible in use, but offer faster restore times and increased
potential for automation.
Disk-based backup can also be done with an ATA
RAID (Redundant Array of Inexpensive Disks) configuration--costing about
$9/GB--and with processes like mirrors, clones, and snaps. These types of
disk-based backup offer the fastest restore times, sometimes reviving data
instantaneously.
RAID systems backup data across at least two
drives with varying levels of intricacy, performance, reliability, and
protection. RAID-1 (high reliability) and RAID-5 (reliable and fast) are two of
the most commonly used RAID levels.
MAID systems have recently created an intriguing
alternative to RAID. Short for Massive Array of Idle Disks, MAID setups also use
disk arrays, but, unlike RAID, only write to one drive at a time. While that
drive is in use the rest of the disks remain idle, dramatically reducing power
use (and usage costs).
In addition, backup responsibilities can now be
delegated to outside vendors offering Web-based solutions, otherwise known as
remote backup service, or RBS.
Once you figure out where to put your data,
you'll need to figure out how to get it there, how often that occurs, and the
amount of data being copied, which is where backup software comes in.
When properly configured, backup software can
automate the process, requiring less-intensive monitoring (rather than
time-consuming manual operation), thereby freeing IT personnel for other tasks.
Generally at least one backup is stored onsite
to maximize recovery time, while at least one backup is stored--or
vaulted--offsite to add one more layer of loss prevention.
Key Features
One of the fundamental choices involved in
delineating a backup process is deciding how much of your data is contained in
each backup snapshot.
There are three major degrees of data capture,
which translate to three main types of backups.
- A
full backup contains every single piece of data every
time, regardless of whether or not a file or folder has been modified. Full
backups take a long time to execute and require a lot of storage space.
However, there's never any doubt about getting a wholly accurate restore. In
addition, that restore can be done in one fell swoop. Since everything's
included, there's no picking and choosing involved.
- An
incremental backup copies only files that have changed
since the most recent backup, be it a full backup or a prior incremental.
- A
differential backup copies only files that have
changed since the original full backup.
Most strategies start with periodic full
backups, and then add more frequent incremental or differential backups to
minimize storage requirements, as both require less space than a full backup. If
a restore is necessary, the full backup is added first, with data filled in from
the subsequent incremental or differential backups.
While incremental and differential backups save
space, they can generate inaccuracies upon restore. Due to the nature of the
data snapshots taken in these types of backups, files can be restored in
multiple versions and/or multiple locations, even after they were purposely
deleted. This can necessitate a manual clean-up that adds to the overall restore
time.
What's
Available
Whether you're setting up the technology
in-house or hiring a service to handle backup for you, the available options are
now more diverse than ever. Costs can vary significantly depending on your
needs. The highest-end products can cost as much as a luxury car, with the
lowest end about equal to a monthly payment. Many fall in the $1,000-$3,500
range.
- Industry leader
EMC offers a full range of hardware and software
solutions, including its popular, high-end
Symmetrix system and
an array of software tools: AlphaStore, EMC Legato NetWorker, Replication
Manager, and EMC Dantz Retrospect (engineered specifically for small and
medium-sized businesses). Retrospect uses a patented progressive backup system
designed to maximize speed while eliminating flawed restores. Retrospect Small
Business Server Edition, for example, begins at $499.
- Veritas Software is
the preeminent vendor of backup software and in the past has occupied the top
two spots on backup software bestseller lists. Veritas offers solutions for
just about every conceivable need with price ranges to match, including the
highly scalable NetBackup and the acclaimed Backup Exec.
- Synology
appeals to SMB customers with its DS-101 DiskStation, a backup storage
solution starting at under $250.
- Data Domain made an
award-winning breakthrough when the company unveiled its DD200 Restorer. Using
proprietary global compression technology, the DD200 is an onsite disk-based
solution that achieves file recovery times orders of magnitude faster than
seen with tape automation, yet remains comparable in cost, coming in at about
$1.95/GB. The Restorer can work in conjunction with the Data Domain Replicator
and your existing backup software to ease bandwidth strain, enabling
cost-effective backup and centralized vaulting across a WAN.
- Overland Storage
offers award-winning options, including the disk-based REO Series and
tape-based NEO Series. The NEO 8000 flagship boasts the industry's most
capacity per square foot of any enterprise library.
- COPAN Systems uses a
MAID configuration in its groundbreaking Revolution 200T, a disk-based virtual
tape library with the tape-like price of $3.50/GB.
- Software
possibilities include platform-specific programs like NovaBackup, designed to
work with Windows NT/2000/XP/2003. Wider compatibility is offered by Syncsort
Backup Express. IPReplicator from Linkpro Technologies works well across
networks. If it's maximum scalability down the road that you're after,
Yosemite Backup Advanced is platform and OS agnostic.
- Remote backup
service providers archive data via the Web, handling all aspects of a backup
operation, allowing smaller companies to reap the benefits of Fortune
500-level service. Specialized online RBS providers include LiveVault, EVault,
and IBbackup. Plans begin under $30/month and stretch toward $800 a month
according to scope.
- Traditional
large-scale data-service providers like IBM, EDS, and HP are beginning to
offer remote backup service as well. Overseas providers Wipro and TCS are
among market leaders on a global level.
How to Choose
Achieving cost-effectiveness for data backup
comes down to three main considerations.
- How
often do backups need to occur?
- How
much data do you need to backup?
- How
soon does your data need to get back online?
Determining your priorities will go a long way
toward determining which solution fits your needs. Some other things you may
want to consider:
- It
Should Burn at Both Ends: Keep in mind the speed at
which your backup system works on both ends of the process. Getting data
quickly back online is crucial. Don't forget to consider how long it takes to
create the backup in the first place.
- Remote
Control: Centralizing your backup operation
eliminates the need to vault backup tapes from remote locations, which can
save hassle and eliminate dependence on remote parties. But, it also eats
bandwidth, which can drive up cost.
- Put
a Scan in the Plan: If you're using incremental or
differential backups, the most efficient software should check for changes
before creating a restore point, otherwise you may end up restoring redundant,
removed, outdated, or even corrupted versions of the same
file.
- Sweat
the Small Stuff: Data floating around on company
work stations and laptops can be just as valuable as information stored in
your network's core. When formulating a backup plan for the hub of your data
universe, don't neglect the data on its user-created edges.