English English

ARTICLES >>> "How much should backup and a copy cost?"

The axiom: if data damaged restoring from backup is the surest way to restore.

Let’s begin with a tip most backup software developers usually forget to remind their potential user. Developers can not be blamed for this because they are professionals and they obviously backup main rule:
Do not store data and backup in one medium or a drive!

If one physical computer hard drive split into two logical drives, for example, C: for data, and D: for backup, it means that the main rule broken! This is still one drive! When hard drive fails as a device to gain access to data of both logical drives equally difficult. Modern hard drives could have different types of faults:
- Magnetic surface destroyed
- Magnetic heads destroyed
- Spindle jammed
- Controller is out of order
- Firmware does not work (or does not work properly)
- S.M.A.R.T. hides a drive for PC.

Two last faults are very interesting because a user will be noticing such faults only with the next computer turn on! Restoring the firmware will be the only a way to gain access to the data. Modern PC motherboards do not support S.M.A.R.T. turn off function (*). The likelihood to restore the firmware yourself is close to zero, only an expert will be able to help. Despite the fact that this type of repair less expensive, it would cost some money (at least $1000) and two days of waiting (**). Search for a software specifically designed for the restoring of lost data from the drive in such a situation is futile and often dangerous. Software of this kind are working more or less intact with the drive that is available at a device level. They allow the restore data affected by computer virus or user error, for example. A program will never restore a firmware as well as never fix defective magnetic heads ...

Assume broken HDD still has a warranty (which is likely since the HDD warranty is up to 36 months and the HDD worked 1-24 months before the first breakage). If so let's replace broken drive with a new one, having spent a little time and no money penny!
Sounds good, but there are problems:
- vendor does not restore damaged data but only produces replacement of faulty devices;
- restoring data after device replacing is impossible.

Suppose there is either no backup or the "main rule" broken and the hard drive is out of order and the price of the data more than repair cost. For example, it might be just a file of 1000 debtors with total amount of $1 000 000. There is only a way; it is necessary to look for an expert who can really help.
The recommendations following which enlarges the chances for the drive to return the lost data:
- do not try to recover a data yourself and do not trust the case to your friends
- do not plug the drive to another computer, the operating system of which will surely try to correct errors on the drive in its favor, but forgot about the user data
- do not run scandisk for the same reason, it will overwrite the file with a list of debtors with some "important" system file. By the way, in Windows XP the "wonderful" scandisk does not support "undo"
- do not trust to "Automatic System Restore" in the same Windows XP, it is a way of reverting incorrect installation of drivers and programs, not involving care of the user information. It is contravening the "main backup rule" also
- be prepared for loss of time and money (minimum 2 days, cost of new HDD plus at least $1000 for a job)
- do not hope that the recovered data get back in the form of operating system (Windows, for example) plus user data. Rather it will be a CD with the files that have been marked for recovery during your first meeting with the expert
- be prepared that time, drive, data and money can be lost simultaneously.

A user once confronted with the data recovery procedure knows well about backup and does not believe in the hard drives reliability. Let’s try to find an ideal backup method based on the "main rule" for ordinary PC user.

First important point is the freshness or validity of the backup. For example, two-weeks-old copy of a project makes sense only when no pity to lose another two weeks to repeat once already made. Another example, a copy of bookkeeping shouldn’t require any comments. It is also clear that a data just before computer shutdown (turn-off) is the ideal source for the backup.
Conclusion: a copy of the data has to be so "fresh" how much time and resources a user is willing to spend to restore lost data release.
Ideal: the correct backup must include the most recent changes and must be made before PC shutdown.

It is so simple to make a backup at first glance. Open Explorer, then choose a file or a folder, copy-paste and patiently wait for the end of the copying process. There is another way, run a backup program and wait for the copying process finish.
Conclusion: a user should remember to do something or should do something to process backup.
Ideal: the user shouldn’t do something and spend a time for the backup process. User never forgets to shutdown a PC, so backup process should be tied with PC shutdown and user log-off.

Note: expecting comments like "there is a number of software that process backups by schedule, etc.," want to say the following:
- any “backup by schedule” software surely reduces performance and stability of the computer;
- changes in project may not be reflected in the backup just because the computer is turned off before the scheduled time;
- changes in project reflected in the backup before the changes were tested, so a rollback is not possible;
- a project is opened with an application. What is copying?

The next important point, how much time the user is willing to spend on backups? The answer is simple, as small as possible! It hard to imagine a user who patiently copying data to CD every night for 30 minutes or making a complete image of the hard drive. The situation can be found for sure, but we are looking for the ideal backup! Storing data on the CD is not the best choice. The reasons are simple. As a minimum, there are low copying speed, poor quality of cheap discs, or the high cost of good discs. With regard to creating an image of HDD, it must be said that this mode of backup requires extremely intelligent use. The drive image is interesting in the first place when it contains an intact and well-functioning operating system, that may take oodles of time and that does not change as often as the user data. Does everyone need this mode of backup that requires considerable time and additional software? No, for certain. If a user uses MS Office only, it is easier to spend an hour for Windows and Office reinstallation after the collapse of HDD because restoring data from a drive image will take just a bit smaller time. The time might be determined by the rate of ~ 400-800 MB per minute.

If you do a comparative analysis of options for storing information on the criteria for storage durability, speed, capacity and cost, strangely enough, it’ll be unreliable HDD. It’s simple: warranty - 36 months, copying speed – more than 40 Mb/sec. capacity - 160 GB, for $80! If you convert the money spent on 365 days and 3-year warranty each daily backup before shutdown would cost about 7 cents! And those payments for virtually full safety of your data!

Now the question arises what to do with HDD unreliability?
The answer is most simple! Nothing! Let HDD producers fight for the quality and the consumer will buy the drives with the highest term of the warranty and exchange HDD for new devices at first failure. The trick that two drives even in a single computer will never go down simultaneously (in the absence of force majeure, of course, like fire or the cheapest power source in the system unit . . .). This secret, by the way, is used in RAIDs. If the backup is on a network or removable drive the probability of force majeure further declines.

Well, it turns out that ideal backup should be done:
- as fast as possible - (differential backup)
- before PC shutdown (turn-off or log-off)
- without user participation
- to an additional drive (hard, network or removable)
- typically, user does not forget to turn off the computer
- user usually does not have a time to process backups
- computers before shutdown keep the most recent changes in files and documents
- two HDD almost never go down at the same time, especially if one of them is external, USB- drive, for example
- users know about hard disk faultiness on the next day usually
- copying to another disk is one of the most rapid and inexpensive way to backup data
- many users already use USB-drives and they do not have to spend money to purchase additional devices
- it is very convenient to keep your data with latest changes with you on USB-drive
- some users are forced to work on different computers with the same data.

This backup concept (and more) the author of the article realized in his "Smart Turn Off" shutdown automation software.

To summarize once said should give an example that shows the correct data location on two hard drives.

2 HDD in system unit

Comments to the picture
1. If the disk 1 collapsed. The disk 1 should be replaced by new HDD. Operating system should be restored on new HDD from the image file from disk 2 (or is reinstalled) and the user data should be restored from the backup, which is on disk 2 using "Smart Turn Off".
2. If the disk 2 collapsed. The disk 2 should be replaced by new HDD, the image file on disk 1 should be copied to new HDD. New backup of the user data will be  created by "Smart Turn Off" automatically.

Best regards,
Iouri Vassiliev.

P.S. Unfortunately, not everyone can afford an additional drive using. Believe the matter of time!
I ought to say a couple of words to alleviate the suffering of the users using a single drive and attitude relevant to their data.

Before or during installation of Windows hard disk should be split into two logical drives. Leave about a third HDD capacity for the operating system (Windows) (drive C: usually) and the remaining capacity use for logical drive "D:". After installation move "My Documents" folder to logical drive "D:" (use menu "Properties" of the folder), and all other important data store on the same drive.
The meaning of these manipulations as follows:
- The operating system will be located in the area with the highest density of data recording, working part-time indicator of the drive health, and important data will be guaranteed shifted to the area with an average density recording, which is the best zone;
- Automatic system recovery and scandisk have no chance to rewrite user data because they have different logical drive location.
With regard to automatic system recovery and scandisk try to make a backup of your data before they start, because no one can guarantee the drive efficiency following the end of the recovery process.

* - it is possible to gain access to HDD at home in this case (to turn HDD S.M.A.R.T. off for a time) but this is my next article material. Registered Smart Turn Off software users are welcome with the question via CONTACTS.

** - regular expert spends about 20 minutes for such HDD fault fixing and you pay for the seldom knowledge and for an expert experience first of all. Two days waiting is something like argument for you to loose your money easily and stay happy.