RAID 10 is a stripe of mirrors, meaning it is multiple drives arrayed in such a way to function as a stripe by distributing parity bits across several drives, and it is a mirror in that each striped disk has a double in which all of the data is copied exactly as well.

For example, RAID 10 would usually be arrayed in a fashion so that you have six hard drives. Each of these hard drives has data striped across three of those drives, and then the other three drives merely clone that data in a mirror.

The Difficulties of RAID 10 and Associated Costs

This format is quite expensive to build and maintain, requiring both several hard drives to operate, as well as very competent software based or hardware based RAID controllers to operate effectively. However, cost aside, this system is incredibly reliable and resilient against data failure.  Linux has an excellent controller called the Linux Kernel RAID 10 to oversee the successful operation of such an array.

In a RAID 10 configuration, such as the one just described, it would be possible for three of the six drives to spontaneously fail at once, and the data that is allocated along these drives should still be wholly intact. While this means that performance could also be increased in hard drive reading and writing applications, the greatest asset in this scenario is the reliability.

A disadvantage to RAID 10 systems in a single computer would be the high energy and heat output, which when dealing with six or more hard drives can cause considerable heat to be expended. In addition, this is a large tax on even highly rated power supplies. If personal computing end-users hope to effectively utilize RAID 10, they must be aware of the demands this system has and prepare their systems accordingly. Those failing drives however must then be replaced, as the remaining drives are then all single points of failure which would destroy all of the data if even one of them was lost. As such, maintaining drives in this or any RAID array is paramount.

High Capacity and Speedy Writing Times

In a databases with  a large amount of data and frequent access, RAID 10 is an excellent choice that allows for quicker write speeds, as it lacks the calculating parity that makes other RAIDS with parity slower.

RAID 10 is a great choice for large-scale computing operations both due to its cost and its great efficiency. Personal computing enthusiasts have also used this configuration’s many benefits, but it is often deemed a complicated and expensive system to those who are still learning their bits from their bytes.

As you may expect, much like RAID 5, RAID 6 is a kind of Redundant Array of Inexpensive Disks in which a computer uses block-level striping. Data is written in blocks on each of the drives on the array. In essence, it leaves parity data across all of the disks, allowing it to rebuild missing sectors of a file.

RAID 6 is similar in design to RAID 5 but with an additional parity block in which files are striped across disks.  RAID 6 is designed to operate under optimal conditions with larger amounts of drives, as it is relatively space inefficient when used in small numbers. RAID 6 is contingent on mathematical computations known as syndromes in order to operate at full efficiency. It is because of these highly important computations that you acquire a card or controller which is expressly rated for RAID 6. This controller does not have to spend time, and lag performance, simply to operate as intended.

The Importance of Implementing the Correct Controller

RAID 6 is controlled by a typical RAID controller in either a software or hardware format, but it requires more precise computing in order to effectively manage the data. With additional calculations called syndromes, RAID 6 can then be configured to run on any number of disks, and provide as much data redundancy as conceivably needed with only additional calculations. These calculations can easily be managed by any specific RAID 5 or 6 controller.

The second layer of parity distribution provides a high level of reliability and more easily recreates files lost after a disk failure when compared to other forms of RAID. The additional data layer of RAID 6 protects against data failure by keeping enough parity bytes on-hand to reconstruct more than one drive, which hedges bets against the dreaded double hard disk failure. This type of failure can wipe out any other RAID arrays effectiveness almost immediately.

Not as User-friendly

This ability to rewrite and rebuild bad data sectors in the presence of two failed hard disks is what makes RAID 6 such an attractive option for both personal and business end users alike. The number of disks involved usually restricts RAID 6 to computing experts or business users, but it is certainly a viable and reliable form of RAID for anyone who has the money to finance such an operation.

The reliability of RAID 6 is what makes it a penultimate form of backup for those who find data integrity to be their chief concern above all. Capacity or storage limits are not as robust as they are with forms of RAID which provide no kind of data redundancy, but you would be hard-pressed to find another RAID array that can protect against catastrophic loss of data nearly as well.

RAID 5 is usually achieved through hardware, such as a high-quality RAID controller. However, with Windows Operating Systems and Linux, they can enact software based controllers to monitor this configuration.
The Data on a RAID 5 array is written in parity stripes, which are then distributed across the drives in the array. These stripes of data are not read every time the hard drive performs a data read, but instead read when a system check finds what is called a cyclic redundancy check error, indicating a failure of the data distribution at some point in the array.

The Affordable Means of High-Level Redundancy

RAID 5 is advantageous to other forms of RAID for its affordable-cost redundancy, which would allow a user to create backups of all their information without sacrificing an entire disk to back up the data of another entire disk.

Four 500GB disks in a RAID 1 configuration for example would ultimately yield 1000GB of space to be filled with data, and 1000GB dedicated to mirroring that data. On a RAID 5 system, that same amount of hard disk space would become 1500GB of usable space instead.

Factors to Consider When Building a RAID array

Although in the grand scheme of things a RAID 5 system is affordable, it is not without cost. At a minimum of three disks, this will be a more costly solution. With that said, although the cost will justify itself in usable space and excellent reliability, it is still a factor for today’s user to consider. It is possible to use a “degraded” RAID 5 array, in which 3 disks exist on the drive, yet one of them is typically inactive until failure or need for backup. However, this would compromise the parity data’s integrity, and thus, the system’s overall reliability.

The parity data can then be used to reconstruct the error and fix the data issue without the main computer ever being the wiser as to what kind of error might exist. If a drive fails, the disks will then reconfigure their data in such a way as to reconstruct the missing pieces from parity files.

When the computer is rebuilding files in case of a failed drive, this is known as Data Recovery Mode. Again, the computer will recognize the failure and prompt the user to take action to recreate the integrity of the RAID array. This failure will not cause the system or programs to run unstably or to cause the system’s performance to plummet, operation will continue at more or less a normal level.

A RAID 5 Tip
Several RAID enthusiasts will always suggest that a user never use hard disks from the same production lot, thus decreasing the odds that a mechanical failure or compromised assembly will strike down all of your hard disks.

Why the Costs of OSB Might be worth it

These offsite repositories are often guarded with the utmost care, and access is usually not freely given to any person who can access data in the main data storage area. This data is guarded by programs and firewalls and many other forms of both physical and electronic prevention to ensure the survival of these vital data vaults.
Offsite data backup allows for one place in which all of the data can be managed easily and by fewer people than if it were dispersed around an office or business. This offsite warehousing of data would be crucial in the event of some kind of calamity in which the data in the main area can be compromised or destroyed.
Remote back up also allows for the creation of a system apart from a main network or mainframe. This allows work to be done on it while other users are still free to operate the main data network, without any need to slow down or wait for the repairs to be completed.

Many companies outsource the maintenance of this data to third party suppliers who will usually backup data on a device much more durable than a hard disk drive, such as a tape drive, or on a server in a process known as e-vaulting.

This off-site collection of data is safe from physical harm, for example if a tornado should level an office building.  Likewise, this vaulted data is safe from electronic harm caused by power surges, outages, or pesky hackers.

It is not uncommon for companies to have hard drive backups of their information, and then another copy of a tape stored in an offsite location.  If one fails, the other still exists in a more accessible form.

Managers and higher level employees can trust that vital documentation is safe even from their own employees, and this gives them another reason to keep their files in a location that is not part of the technological network that comprises their data.

Off-Site Backup is Not Right for Everyone

Offsite backup is a laborious process, involving the coordination of many devices, people and information. It is for this reason that it is largely impractical for many personal users, although not completely unheard of.
Remote backup is reliable and imminently stable, even if it is not always inexpensive. If theoretically an electro-magnetic pulse were to rip through Honda’s blueprint database, it would still be far less expensive to retrieve the data from magnetic tape drives than it would be to commission engineers and designers to completely redesign the companies’ models.

Hard drives must spin at 5,400, 7,200, 10,000 or even more Rotations Per Minute (RPM) in order to operate. It is this high-capacity spinning that allows for what is known as “random access” meaning the hard drive is easily accessed at any point from any point. The high-speed rotations however also significantly shorten the life-span of the machine. It is very common for the needle of a hard drive (much like the needle on a record player) to skip or become unbalanced, causing the hard drive to fail.

The Long Shelf Life of Tape Drives

The tape drive is more akin to a cassette in its durability. The spool of magnetic tape must be protected at all times, but as long as the integrity of the tape is assured, the tape drive will likely have a much longer life, all things equal.

The tape drive is the paradigm of the old adage “slow and steady wins the race”. A tape drive has write speeds that rival quicker hard drives, even though its read-time might lack, as the spool of tape must be rewound and sped up to an exact point. It is still a distinct advantage over the long haul to keep important data on a more stable machine, than on one which is faster.

Tape drives are much more durable than media, which is another popular option for backing up data. CDs and DVDS have high capacity, but then require peripheral devices to read the data, and they too must be kept in pristine condition, otherwise they may very well be rendered useless by a few scratches or drops. And of course media can only hold so much space, which is usually not nearly up to the level of modern tape drives (please disregard this statement if you are still employing your 100MB tape device from 1977).

The Longer Times Associated with Tape Drives

Tape drives are the long-term money and time saver, even though they may be more taxing on the patience in the short term. Tape drives do not require the immense amount of power from your system, saving energy and heat from potentially ravaging a system. Tape drives do not require defragmentation or sector checks, and most viruses are built to prey upon the hard drive, and not the tape drive, thus adding huge amounts of safety and reliability.

Most tape drives are modular, which means they can be portable should the need arise. While external hard drives also exist as an option, portability is standard for nearly all tape drives made in modern times.

The reliability and low cost of the tape drive is well worth the investment to the user who must keep their data immaculate and easily accessible at all times, depending on your application, these devices may prove themselves to be anything but obsolete.

The Right Configuration for the Job?

RAID 1 provides a maximum level of data redundancy, but it does not include as many appreciable speed increases to a system as you would find in a RAID 0 array. Instead, RAID 1 is the choice of the user who above all needs to have safe copies of their data at hand and protected from the errors and failures of a hard drive.

In terms of speed, once a RAID 1 array is set-up, it is possible for a computer to read two different portions of the same data at the same time, which would in theory double the existing speed of the system in a data management term known as duplexing.

This speed boost is highly contingent on the manner of hard disk used, as well as the RAID controller involved in the situation. An older IDE configuration can still only read a single disk at a time in trying to access information. Therefore, RAID 1 is a slower alternative with these older drives.

A RAID 1 array can only be as big as the smallest disk in the configuration. If you were to array two hard disks, one of 100GB and another of 250GB, your overall RAID array would then be 200GB in size, with 150GB of hard drive space still free.

The biggest disadvantage of RAID 1 is that it effectively halves the possible storage space one would have on their computer. Instead of having two 250GB hard disks, used independently for 500GB of space, you would have effectively half of that to fill with data. The other half is dedicated to maintaining a mirror of that data instead of additional space for storage.

The Stability of RAID 1 Arrays

Reliability is quite high if you were to assume that a hard disk would be replaced in the event of a failure or malfunction. Assuming this, there is a .0004 percent chance that both drives will fail and your data would be lost in its entirety. Breaking these kinds of mirrors would be several years of bad luck indeed at that level of reliability, but assuming your disks have an adequate RAID controller and are arrayed correctly, you should be quite safe from the aggravating black hole of data loss.

Still the administrative advantages of RAID 1 make it a popular option for those looking to keep the integrity of their data intact at all times. A mirror can be split, allowing one disk to be removed and backed up or replaced, and the system will not miss a beat. Instead it will continue to function normally, and then when the hard disk is reactivated or reinstalled after a failure, rebuilding the mirror requires minimal effort.

RAID 0 uses striped volumes, which are also called a stripe set in order to backup data in such a way as to partition it between drives equally.

Advantages

RAID 0 is different from other RAIDS in that it provides no data redundancy, which means there are not replica files backed up onto another drive, rather, they are split between drives in alternating volumes. File A for example, would be separated into four pieces, with pieces 1 and 3 on one volume of the array, and 2 and 4 on the other.

A RAID 0 array allows for better performance as the hard disks are breaking up files and saving them in pieces, rather than processing whole files, as many times as you have disks in an array. This lack of complete file duplication is what is known as no redundancy.

In the event of a hard disk failure, it would be easy for the files to reconstitute the lost data with the use of parity bits, or pieces of data that can recreate files based on only pieces of existing data. This is as opposed to other RAID arrays, which simply make duplicates of the entire file.

Disadvantages

RAID 0 best operates when utilizing hard disks of the same size. If someone were to configure two hard disks of differing size in a RAID 0 array, they would find that they would only be able to array disks to the amount of the smallest disk. Still, it is vital to keep in mind in any RAID array, the more hard disks that are being utilized in any system, the probability of a system failure increases. Thus, it is important to strike a balance between reliability and increased performance from RAID arrays.

For example, if you had a 100GB hard disk, and a 250 GB hard disk, a RAID 0 array would only be 200 GB, with one drive being used completely, and the 250GB hard disk using only a portion of its’ space. RAID 0 is best used in large-scale computing or business operations, in which redundancy or maintenance of files is less important than the actual speed at which the data is stored.

RAID 0 was not a usage of the configuration as originally envisioned by the inventor of the RAID system. Still, RAID 0 is no more unstable than other RAID configurations, which is a desirable facet for business applications indeed.

RAID 0 can only support as many drives as there are letters in Windows Operating systems. Also, in gaming applications, RAID 0 will provide minimal gains on performance. The RAID controller will determine the speed of transfers between disks on a RAID 0 array, and so it is vital to ensure a system has the best possible controller in order to maximize RAID 0 speed boosts.

SiteVault allows anyone to backup their files via FTP and MySQL databases significantly easier than before. There is very little to say about SiteVault as it can be summed up as a tool to backup your sites to your own PC, however, here’s the feature list from the official site to give you an idea of what it actually does:

• Manage multiple accounts to organize your backups
• Backup your data (files and MySQL databases)
• Restore your data
• Incremental backup (only modified files are downloaded to avoid downloading the same file twice)
• Easily cleanup unneeded backup files
• Handle virtually unlimited FTP requests
• Doesn’t matter if your host does not allow outside connections to the SQL server

The verdict? If you’re a Windows user, have multiple sites, and want an effective way to back them up then definitely check out Site-Vault.com, many people swear by it and it has a great set of features for a very nice price.

Some facts…

• The software is $99 but there is a free trial to take advantage of before you buy
• A selection of video tutorials are available that run through all the options
• Runs on Windows 2000, Windows XP & Windows Vista

Okay, so what does this do?

DriveImage XML allows you to create bootable image files of your partitions and drives. As an example, say your system decided one day it wanted to die, you’d have one of these image files on a disk. Just insert the disk, reboot the computer, run the software from the disk, and your computer will be fully restored. Yep, it’s that simple. Okay, if you didn’t understand that paragraph then DriveImage XML might not be for you as it’s not a “push this button and all will be fine” type of application.

Why use DriveImage XML?

…and of course, it’s completely free! Yep, that’s the real kicker, this application, which what most consider of commercial-quality, is complete freeware. The only downside is no support is given, but the target audience for this application is computer-literate users wanting advanced backup methods anyways.

But that’s not all! What’s also great about this software is all backups our outputted in XML format. So?…well that means you can use a number of 3rd party tools to restore your data which is nothing to complain about.

Currently DriveImage XML runs on…

• Window Server 2003
• Windows XP
• Windows Vista

Overall?

If you’re a Windows user, are well-versed in the ways of the computer, and want a complete, reliable backup solution then creating bootable disk images with DriveImage XML is definitely a strong recommendation.

Why would you want to retain backups?

• Data Corruption. For a variety of reasons backups can be corrupted in some way. If all your backups are corrupted, then you have a major issue, but what if every backup is fine until one day the system decides something is wrong and an entire backup is botched. If your backup is automated and you don’t read error logs you may not realize there is a problem. Now, if you haven’t retained previous backups and something goes wrong, guess what? You’re stuck with a corrupt backup that is no use to anyone. If you have older backups however you can take that data, and if you’re lucky salvage the newer data from the most recent backup.
• Hackers. Yes, the “H” word. It’s no secret that certain people in the world like to get into computer systems for purely malicious reasons. If a hacker was to access your computer system or web server they may destroy your data, and then after they’ve done that a backup of the destroyed data may be automatically generated if that’s what you’ve set it to do. Great, now you have a backup of destroyed data. Kind of wish you retained your backups now, right?

There’s no real excuse not to retain backups with continual lowering costs of data storage, and more compact systems/hardware. Also, unfortunately there is no hard and fast rule on how long you should retain backups for, it really depends on the purpose of, and how much data is being fed through your system.