Wanted: Backup Strategy

I have a machine with five disk drives:

(C) is for Window 7 + software: 150GB (10,000 rpm: sweet)
(D) is to backup (C): 150GB (just like C)
(E) is for data: 1TB
(F) is to backup (E) 1TB (just like E)
(G) is for paging and spooling (Photoshop, etc): 380GB

My best case scenero is that I'd like for (D) to duplicate (C) and be bootable in the event something goes wrong with (C). Also, and the end of the day I'd like to run a batch procedure to backup what has been changed or added to (E) to (F), and also in that process backup my Microsoft Outlook files.

I've tried RAID-1 for two years and have given up. The RAID arrays were broken too many times, and I just have to get on with my life (LOL). So I'm in the process of rebuilding my system without RAID.

Suggestions appreciated !!
20 answers Last reply
More about wanted backup strategy
  1. You don't really have a backup if you're just copying files to another drive in your system. And that goes doubly if the "drive letter" that you're copying to is just another partition on the same disk.

    There are too many things that can go wrong that would affect both copies of your files, from corruption to power hits to theft of the entire system. You're not well protected unless you copy your files to an external drive which you then store separately. Ideally you'd have two offline copies, one of which is offsite.
  2. sminlal said:
    You don't really have a backup if you're just copying files to another drive in your system. And that goes doubly if the "drive letter" that you're copying to is just another partition on the same disk.

    These are five (5) separate drives. None are partitioned.
  3. Regardless having the important Data on a separate external is a very good idea. I personally only back up (two copies) that way and only user files!
  4. Well, if all of the HDDs are up and running, during a power failure, there is a possibility that all HDDs be affected altough it is highly unlikly for it to happen to all of your drives.

    My suggestion is to get two USB adapters for two of your drives, take the HDDs outside (the ones that will be used for backing up) and connect them through those adapters (USB HDD adapters may need external power supply, that are included in the kit) only when you want to back up anything, they should not be running at the same time with the source of the the data, as we need those back-up HDDs to be less used.

    http://www.newegg.com/Product/Product.aspx?Item=N82E16812119279

    Make sure it supports up to 1 Tb HDDs.
  5. websquad said:
    I have a machine with five disk drives:

    (C) is for Window 7 + software: 150GB (10,000 rpm: sweet)
    (D) is to backup (C): 150GB (just like C)
    (E) is for data: 1TB
    (F) is to backup (E) 1TB (just like E)
    (G) is for paging and spooling (Photoshop, etc): 380GB

    My best case scenero is that I'd like for (D) to duplicate (C) and be bootable in the event something goes wrong with (C). Also, and the end of the day I'd like to run a batch procedure to backup what has been changed or added to (E) to (F), and also in that process backup my Microsoft Outlook files.

    I've tried RAID-1 for two years and have given up. The RAID arrays were broken too many times, and I just have to get on with my life (LOL). So I'm in the process of rebuilding my system without RAID.

    Suggestions appreciated !!



    Hey just wondering, why did your RAID-1 keep breaking? Isn't RAID-1 suppose to increase data safety? (Never used raid before)

    And could something like RSYNC or GHOST work for you, saving image copies to an external drive via USB?
    http://buy-static.norton.com/norton/ps/1up_ca_en_gh.html?om_sem_cid=hho_sem_sy:ca:ggl:en:e|kw0000011946|10365111201&country=CA
  6. q4quality said:
    Hey just wondering, why did your RAID-1 keep breaking? Isn't RAID-1 suppose to increase data safety? (Never used raid before) And could something like RSYNC or GHOST work for you, saving image copies to an external drive via USB?
    http://buy-static.norton.com/norton/ps/1up_ca_en_gh.html?om_sem_cid=hho_sem_sy:ca:ggl:en:e|kw0000011946|10365111201&country=CA

    Well, I've heard the same things you have about RAID-1: it is supposed to be a duplicate copy of the master drive onto the secondary drive, with everything done automatically. I had a pair of 150GB (alas, the marketing term for disk size ... it is really smaller) 10,000 RPM Western Digital drives for my system/software, and a pair of 1TB Seagate drives for my data (I run about 20 websites). It seems like every 6-8 months something would go wrong and I'd spend a lot of time getting drives replaced or re-synching an RAID array. I just got tired of all of that.

    My solution based on responses from this forum, a couple of other forums, and an inquiry to Western Digital support:

    (1) Downloaded Acronis True Image from Western Digital Tech Support. It was easy to install and did a full wipe of my drive "D" and then a system image copy of my "C" drive to "D" if just a few minutes. Couldn't believe how fast it was.

    (2) Tonight I'll build a batch command file to back up all my data ("E") to my data backup ("F") with a mouse click ... will include the Microsoft Outlook data files as part of the batch backup, and will run that every night. Will run Acronis once every week, and after every major software addition (alas, I still have a lot of software to add to "C").

    Thanks for your help and ideas!

    PS: You 'gotta love this forum software. I have not posted very much on this forum, so it classifies me as a "youngster" .... alas, I wrote my first FORTRAN program in 1962, wrote my first spreadsheet (also in FORTRAN) in 1966, built my first home computer (Heath H8) in 1977, four years before the IBM PC ... made a decent living in IT and am now happily retired doing small websites for non-profits on a pro bono basis. Life is sweet at 71!
  7. Lol... good luck and let us know how it goes in a few months, Youngster :D


    P.S. There was even a lawsuit against WD and others over their scammning of the consumers. In the end they didn't get affected much though, but do have to tell the consumer now about there $$ saving, marketing defination of KB / MB / GB.

    http://www.foxnews.com/story/0,2933,201269,00.html
    http://en.wikipedia.org/wiki/Binary_prefix#Standardization_of_dual_definitions

    Shame they just weren't forced to adopt the binary values, because addressing in the computer is not base 10.
  8. q4quality said:
    Shame they just weren't forced to adopt the binary values, because addressing in the computer is not base 10.
    Here I disagree. Memory sizes are always powers of 2 because of the way memory addressing works - but the internal addressing of disk sectors uses a cylinder / head /sector number which has nothing to do with powers of 2. There is basically zero relationship between how much stuff you can cram onto a disk and the binary numbering system, so there's no good reason to base disk capacities on that system.

    If anything, I think Microsoft is to blame because Windows Explorer insists on displaying sizes using binary "KB", "MB" and "GB" suffixes. The problem with this is that for the uninitiated user they're not consistent - 1000KB doesn't mean the same thing as 1MB. And since transfer rates are always reported using decimal multipliers (even by the most techie of industry insiders), it means that there's no consistency between transfer rates, reported file sizes, and transfer times.

    For a company that supposedly prides itself on an easy-to-understand user interface, it's ludicrous IMHO.

    It seems that techies often disagree with me on this point (and few people are more "techie" than I), but I think the acid test is this: If Windows Explorer had a "report file and disk sizes in decimal or binary" option, which would you choose? I have little doubt that the vast majority of people would choose decimal.
  9. Interesting points. But my turn to disagree.

    The addressing information that gets passed to the drive is base 2 as it comes from an address bus off the computer. By not conforming to base 2 math, disk manufactures can play marketing number tricks in-order to save money by not providing a full base 2 amount of storage (i.e. 1024^4 = 1 binary TeraByte v.s. (1000)^4 = 1 decimal TeraByte which of course is smaller.).

    Sure you don't *have to* use the full address space of the bus that feeds the disk drive, but people would expect you too as it is a waste not too fully utilize the address space; the only advantage is making a slightly cheaper product with slight less storage. Of course the disk manufactures where playing a numbers game where they made a slightly cheaper product and let us think it was still a full amount of base 2 math storage, and then pocketed the savings (thus the lawsuit).

    Anyway, the addressing space waste argument is even more relevant if we talk about SDD drives or USB flash sticks, and how addressing of flash ic memory cells are done!

    Now I agree the SI prefix K = Kilo general means 1000, but when talking about computers it has been commonly accepted to mean 1024. Anyone working in the electronic hardware design or low level software design industry will know how this is a necessity to simplify there work. No sane hw designer would use K as 1000 instead of 1024!

    Since Bytes (and bits) are computer terms I do not see the confusion in saying KB = 1024 Bytes and MB = 1024*1024 Bytes.
    It is easy to keep things straight:
    1) When talking computer terms numbers are in binary units (K = 1024, M = K*K = 1,048,576, etc...). And
    2) When talking about non-computer terms numbers are in decimal (K=1000, M=1,000,000, etc...)

    There is no confusion that 1000KB != 1MB, since these are computer terms and we all know 1024KB = 1MB. :)
    This is very simple.

    What makes it confusing is when marketing/sales/non-techi management people in the computer industry start stating things in decimal when they should be in binary!

    Now you are probably right that non-techies would choose base 10 reporting if given a choice and the choice pointed out to them (by a prompt during installation say). But by definition, non-techies almost assuredly don't care about that sort of thing and thus if not given a choice would not be confused by the base 2 reporting (and likely pleased when they see the number is larger then they thought it would be). Most techies, on the other hand, would choose base 2 reporting.

    So perhaps a better acid test would be this: Is the terminology consistent from low level designer to high level user? If it is, then designers and end users can talk the same language and no confusion will be had.

    It is a total shame that IEC tried to redefine the terms KB, MB, etc by introducing Ki, Mi, etc.. . KilobeeByte, MegabeeByte..., etc.. lol yah right! Like we aren't intelligent enough to figure out the meaning of KB by its usage.

    Anyway, that's my opinion and one that WD lawyers don't share.

    Sorry for the rant ;)... it was late. Keyword, 'was'.
  10. q4quality said:
    Most techies, on the other hand, would choose base 2 reporting.
    I doubt it. I'm about as techie as they come - I was writing programs in machine language back in the 70's and was writing assembler programs for an embedded microcontroller just a couple years ago. I've worked on various computers where reading core dumps required fluency in binary, hex and octal. I'm very, very familiar with this stuff.

    Yet when I sit in front of a GUI and use a computer, there's absolutely nothing that requires me to deal with file sizes as powers of 2. Nothing. There is no reason to use them, period.

    It's far, far more useful for me not to have to worry about the differences when I'm looking at a file that Explorer reports as containing "986,004KB" and wondering if it will fit on a memory stick that Explorer reports as having "963MB" of free space. Yes, as it turns out it will indeed fit - but it's not obvious, even to a techie. The only way to be sure is to right click on everything to bring up properties dialogues so that you can get both values in decimal and compare them using the same number base. Why on earth would anyone, techies included, prefer to do that?
  11. All your pointing out is the inconsistency in the windows display method for binary number display.

    I still feel that if the terminology is consistent from low level designer to high level user, then designers and end users can talk the same language and no confusion will be had. Its just a bit of learning. We are all taught base 10 math in elementary school. Kids nowadays should be also be taught base 1024 math, if not base 2 math.

    And I too am as techie as they come, being both a low level software designer and an electronic hardware designer.
  12. This conversation is very interesting, especially the 963 mb of free space example. :)
  13. q4quality said:
    All your pointing out is the inconsistency in the windows display method for binary number display.
    Well, that's the whole point, isn't it? Using 1024 multipliers 1000MB is not equal to 1,000,000KB, so the numbers are inconsistent and not directly comparable. There's only two ways to make them consistent, either eliminate the multipliers altogether and just show the size as a long, comma-separated decimal value so that mismatched multipliers can never arise - or use decimal-based multipliers. Well hey, both those alternatives mean using decimal.

    It doesn't matter how long you've been a techie or how much training people have, it's never going to be obvious whether or not a 986,004KB file will fit onto a drive with 963MB of free space.

    IMHO, saying that Explorer "should" use binary multipliers is akin to saying that Explorer should show file attributes as 0x0023 instead of "AHR" to indicate a file with the archive, hidden and read-only bits set. It's hex, and it's the way the designers coded the thing, right? Shouldn't we all be using the system the way the designers do?

    What's that you say? Too complicated, there's no benefit in showing the hex values versus symbolic codes that are more easily understandable? Well that's exactly my argument for using decimal multipliers instead of binary.
  14. sminlal said:
    "q4quality wrote : All your pointing out is the inconsistency in the windows display method for binary number display."

    There's only two ways to make them consistent, either eliminate the multipliers altogether and just show the size as a long, comma-separated decimal value so that mismatched multipliers can never arise - or use decimal-based multipliers.


    I guess I need to clarify this for you.
    I will repeat a few things, but here I go:

    It is easy to keep things straight:
    1) When talking in computer terms, numbers are in "binary" units (with multipliers K = 1024, M = K*K = 1024*1024, etc...). And,
    2) When talking about non-computer terms numbers are in decimal (with multipliers K=1000, M= 1000*1000, etc...).
    And,
    3) when displaying "binary" numbers, to avoid confusing people that are use to decimal number representation, simply display them in this format: # S (Bytes or whatever)
    - where # is a number from 0 - 1023 (or 1024, if you like), and
    - where S is a multiplier like K=1024 or M=1024*1024 or G, or T, etc... . Powers of 1024, instead of powers of 1000 that SI multiplier notation uses. # is 'basically' a base 1024 number, and well suited to represent a true binary number.

    For example:
    - 1023 B or
    - 1024 B or
    - 1 KB (here #=1 and S=K=1024) or
    - 962.89453125 MB or
    - 963 MB (here #=963 and S=M=1024*1024) or
    - 1000 MB or
    - 1023 MB or
    - 1.0009765625 GB.

    All the above numbers are easy to compare with each other.

    You can even convert 962.89453125 MB to decimal notation without multipliers if you want to (its 1009668096 Bytes), but its pretty obvious that this number is less then 963 MB when displayed correctly.

    This is just an appropriate and consistent method for displaying "binary" numbers to people use to a decimal number system.

    And that is the whole point.
    Well that, and that we should keep a consistent number representation language between the whole chain from low-level designer to end user (i.e. We should keep KB, MB, GB, ... as there "binary" meanings of 1024 B, 1024^2 B, 1024^3 B, ... instead of using the SI notation of 1000 B, 1000^2 B, 1000^3 B, ...).

    What windows does is an inconsistent display of "binary" numbers, and for you this inconsistency is the problem.


    p.s. We are talking about a number representation "language" for numbers used in computers v.s. other areas of life.
    And yes I do think we should keep consistent language between designers and end users, in this respect at least.
    That does not mean that if a designer decides to represent 35 decimal as "AHR" for other purposes, that they should not do so.
    But God forbid a low-level software designer that decides to represent 0x0023 as 35 decimal in the situation you described above.
  15. q4quality said:
    Well that, and that we should keep a consistent number representation language between the whole chain from low-level designer to end user (i.e. We should keep KB, MB, GB, ... as there "binary" meanings of 1024 B, 1024^2 B, 1024^3 B, ... instead of using the SI notation of 1000 B, 1000^2 B, 1000^3 B, ...).
    But why? What justification can there be for using a system that most people are unfamiliar with? What is it about using binary units that is better than using decimal units?

    Engine designers deal in torque and rpm, yet cars still have odometers calibrated in miles or kilometres per hour. PVR (video recorders) devices have hard drives, yet the user interface shows percent or hours free, not bytes. I really can't see any justification for obfuscating information just because an engineer works in different units than the end user.
  16. Pointless to argue about it the whole thing is based on two condition off and on that is 0, 1!
  17. sminlal said:
    But why? What justification can there be for using a system that most people are unfamiliar with? What is it about using binary units that is better than using decimal units?

    Engine designers deal in torque and rpm, yet cars still have odometers calibrated in miles or kilometres per hour. PVR (video recorders) devices have hard drives, yet the user interface shows percent or hours free, not bytes. I really can't see any justification for obfuscating information just because an engineer works in different units than the end user.


    Dude, haven't I already explained this too you with the previous posts?

    The benefit of consistently using 1024 multipliers instead of 1000 multipliers for computer "memory" discussions is less confusion, as not doing so causes confusion and obfuscation to people already schooled in the defacto standard usage of KB = 1024 Bytes. As well, it also causes unnecessary confusion between low-level designers and end users (re: customer technical support).

    Perhaps you are trying to argue that the majority of non-techie users who care about this at all would be confused by KB = 1024 bytes. But I have already pointed out that with a small bit of education about the definition of KB/MB/GB... (fine print say :) ), and shown that with a proper display of the "binary" numbers, this point would be moot. Notice that drive manufactures are currently trying to use small print to "re-define" what KB/MB/GB... mean.

    As for your points on engine designers and PVRs, you are comparing apples to oil drums.
    RPM is in terms of revolutions per minute, not distance. While the "Odometers" are displaying a distance measure. But I am sure you knew that already.
    The proper analogy for your argument would be to argue that odometers should represent distance in miles instead of kilometres (of course it really should be the other way around ;) ).
    What you are pointing out is unit conversion: rpm to miles/kms, bytes to time. This is not the same as our KB = 1024 Bytes v.s. KB = 1000 Bytes argument. We are not doing unit conversion here. Its all bytes.

    Sure there are situations where terminology used by designers are not understood by end-users. BUT when we have the opportunity to not have this situation and instead have consistency in the terminology usage from designer to end-user, I argue that it is a terrible shame not to embrace that. Consistency is the key here and any non-techie would agree that consistency would be better.

    In addition, I argue that the benefits to consistently use 1KB = 1024 Bytes and not 1000 Bytes (with 1 M = 1K * 1K, etc) in this area is more beneficial then to not do so. If you truly are "as techie as they come" you would understand this. Especially since, as I have already mentioned, it takes very little effort: to school the non-techie in the "proper" usage of the term KB (1024 Bytes) and to display these numbers in a way that is easy for a non-techie (or anyone) to compare/understand.


    So short answer is "Why not". You haven't yet given a good reason why not, where as I have countered all your proposed reasons and given you solid reasons "Why to". Well at least IMHO.

    Anyway, at this point it is fairly clear that you will not let [my] logical reasoning sway your views on this matter, for whatever reason. So I would be more then happy to agree to disagree, and you can chalk this up as just another case of a techie disagreeing with you.

    If you would really like to continue this discussion, you can go ahead and create a new thread about this, as we have hijacked websquad's thread long enough (I am kinda surprised a MOD has not locked this thread by now ;) ).

    Have a good Easter sminlal.


    Quote:
    P.S. There was a lawsuit against WD and others over their scammning of the consumers. In the end they didn't get affected much though, but do have to tell the consumer now about there $$ saving, marketing defination of KB / MB / GB.

    http://www.foxnews.com/story/0,2933,201269,00.html
    http://en.wikipedia.org/wiki/Binary_prefix#Standardization_of_dual_definitions
  18. q4quality said:
    The benefit of consistently using 1024 multipliers instead of 1000 multipliers for computer "memory" discussions is less confusion, as not doing so causes confusion and obfuscation to people already schooled in the defacto standard usage of KB = 1024 Bytes. As well, it also causes unnecessary confusion between low-level designers and end users (re: customer technical support).


    I feel that there would be a lot less net confusion in the world using a decimal base for disk capacities since there are far more end users who are in fact confused by binary compared to techies, and the techies and support people are a lot less apt to get confused because they understand the difference between the two systems. In short: the needs of the many outweigh the needs of the few.

    Looks like we'll just have to agree to disagree ;)
  19. sminlal said:
    In short: the needs of the many outweigh the needs of the few. Looks like we'll just have to agree to disagree ;)


    Well as I described above, it is extremely easy to alleviate any end-user confusion with fine print and the proper displaying of "binary" (1KB = 1024 Bytes) numbers. It just takes a small bit of effort for a net benefit to everyone. There would be no confusion.

    But I don't have a star trek quote for you (which I thought was very good usage btw), so lets just say:

    Quote:
    q4qualtiy wrote:
    Sure there are situations where terminology used by designers are not understood by end-users. BUT when we have the opportunity to not have this situation and instead have consistency in the terminology usage from designer to end-user, I argue that it is a terrible shame not to embrace that. Consistency is the key here and any non-techie would agree that consistency would be better.

    In addition, I argue that the benefits to consistently use 1KB = 1024 Bytes and not 1000 Bytes (with 1 M = 1K * 1K, etc) in this area is more beneficial then to not do so. [Especially since using such terminology is a necessity for the designer and] , as I have already mentioned, it would take very little effort to school the non-techie in the "proper" usage of the term KB (1024 Bytes) and to display these numbers in a way that is easy for a non-techie (or anyone) to compare/understand.


    Anyway, sounds good to me.
  20. Please get back on topic and suggest a solution for the PO's question.
Ask a new question

Read More

Hard Drives NAS / RAID Backup Disk Drives Storage Product