ATI vs. FDISR in space

Discussion in 'backup, imaging & disk mgmt' started by ErikAlbert, May 21, 2006.

Thread Status:
Not open for further replies.
  1. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    No, it isn't a SF-movie :D , but an experiment to see
    1. how much space ATI needs to keep 10 images with normal compression.
    2. how much space FDISR needs to keep 10 snapshots on system partition [C:]
    Both started with the same partition [C:] = winXPproSP2 + Softwares, but no personal data.
    My system partition has a maximum capacity of 48,8gb.

    After each backup, I wrote the used space of the system backup partition [E:] down.

    Acronis True Image .........Bytes ........KB ....MB ...GB
    ------------------ -------------- ---------- ------ -----
    System Backup-01.. .3,166,523,392 .3,092,308 .3,020 .2.95
    System Backup-02.. .6,264,217,600 .6,117,400 .5,974 .5.83
    System Backup-03.. .9,362,087,936 .9,142,664 .8,928 .8.72
    System Backup-04.. 12,460,077,056 12,168,044 11,883 11.60
    System Backup-05.. 15,558,234,112 15,193,588 14,837 14.49
    System Backup-06.. 18,656,493,568 18,219,232 17,792 17.38
    System Backup-07.. 21,754,859,520 21,244,980 20,747 20.26
    System Backup-08.. 24,853,340,160 24,270,840 23,702 23.15
    System Backup-09.. 27,951,906,816 27,296,784 26,657 26.03
    System Backup-10.. 31,050,567,680 30,322,820 29,612 28.92

    So ATI needs 28.92gb to store 10 images of 6.44gb
    Each image backup took lesser than 3 minutes.
    -----------------------------------------------------------------------------

    After each reboot I wrote the used space of the system partition [C:] down.
    So I rebooted each time with the last snapshot to create the next snapshot.

    FirstDefense-ISR.. .........Bytes ........KB ....MB ...GB
    ------------------ -------------- ---------- ------ -----
    Without FDISR..... .6,914,281,472 .6,752,228 .6,594 .6.44
    System-Snapshot-01 .6,948,712,448 .6,785,852 .6,627 .6.47
    System-Snapshot-02 11,669,884,928 11,396,372 11,129 10.87
    System-Snapshot-03 16,397,799,424 16,013,476 15,638 15.27
    System-Snapshot-04 21,138,767,872 20,643,328 20,160 19.69
    System-Snapshot-05 25,871,982,592 25,265,608 24,673 24.10
    System-Snapshot-06 30,609,133,568 29,891,732 29,191 28.51
    System-Snapshot-07 35,349,893,120 34,521,380 33,712 32.92
    System-Snapshot-08 40,094,404,608 39,154,692 38,237 37.34
    System-Snapshot-09 44,842,553,344 43,791,556 42,765 41.76
    System-Snapshot-10 49,596,735,488 48,434,312 47,299 46.19

    Was there an 11th snapshot?
    No and it was a big relief for me, because I had only 2.63gb left, which wasn't enough for a next snapshot.
    But there was still the option "<New Archive>" to choose, which I didn't test yet.

    So FDISR needs 46,19gb to store 10 snapshots of 6,44gb
    Each snapshot took 10-12 minutes.
    -------------------------------------------------------------------------
    Conclusion :

    FDISR minus ATI = 17.27gb. So FDISR needs 17.27gb more space than ATI.

    Booting from another snapshot is in FDISR indeed fast, but I lose 10-12 minutes to create a new snapshot and creating a new archive won't be faster.

    Restoring from another image in ATI is slow and requires about the same time to create a FDISR-snapshot.
    But backup in ATI is less than 3 minutes, including verification.

    The big advantage of FDISR is booting from a snapshot.
    Because you normally reboot more than creating a new snapshot/archive you will save alot more time, than restoring an image with ATI.

    --------------------------------------------------------------------------
    This was also an opportunity for me to test the restore of ATI again in an extreme situation.
    I restored my previous backup without FDISR on it over an almost full system partition [C:] with 10 snapshots without any problems and of course after that there was no FDISR anymore.
    I will do some more testing of course.
     
  2. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    Hi Erik

    I don't know if you have your external drive yet. Some idea's to play with for your testing:

    Time to make a new snapshot in FDISR, vs time to make a new archive on an external drive. Then time to make a new snapshot from an external drive archive. For me the archive time is about half of the snapshot time.

    Also another test that might be interesting. This assumes you want a backup on a daily basis and maybe want to keep the last 5. Then build 1 disk snapshot, and 5 external archives. Here the purpose of the disk snapshot would recovery fall back and the archives your backup.

    Once you are setup compare the time, do to and image every day vs the time it takes to refresh your disk snapshot, and one of the archives. In this scenario, you wouldn't need a new image very often.

    I'll be very interested in the results.

    Pete

    PS I've been very incouraged by your results with Acronis. You are using it the way I am and it's nice to know the recovery works well.
     
  3. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    Peter,
    No I don't have my external harddisk yet, but I need to get familiar with FDISR first.
    I use my second internal harddisk as backup for the moment, which wasn't my intention, but I need a backup for all my other experiments and learning new softwares takes time too and will keep me busy for awhile.

    Acronis True Image is working great here and I tried to break it in any possible way I could think of, but whatever I do, it keeps on working without any problem. :D
    I will test ATI with FDISR on my computer now, maybe I can break it this way. :D
     
  4. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    Hey you can do archives on the 2nd drive, just as good. Be sure to turn off the preboot, before imaging with FDISR installed. Just safer that way.
     
  5. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    Yes, I know I can save archives on my 2nd drive, but both harddisks (WD Raptors 10000rpm) are faster than my external harddisk (7200rpm)
    So backup/restore TIME won't be the same.
    If you are interested in the TIME, I will give you the wrong time, when I use my second internal harddisk. :D
     
  6. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    It's still a good measure, because the same is true if you image to the external drives's.

    Hey just out of curiousity are those drives quiet?
     
  7. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    I have a pretty good isolated computer case "Antec Case Sonata II".
    I don't hear anything, but my hearing isn't what it used to be. So I wouldn't count on me. :D
     
  8. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    Peter,
    Creating a snapshot takes 10m26s
    Creating an archive takes 9m12s
    Restoring an archived snapshot takes 8m29s

    The time isn't of course always the same. It depends on the writing speed, which changes constantly, but I don't think I have to tell you this. LOL

    So it seems that working with archived snapshots is indeed faster than working with unarchived snapshots.
    The difference isn't big in this example, but my system partition [C:] doesn't contain many applications either, because this is a test machine for the moment.
    The difference will increase for larger volumes.


    P.S.:
    FDISR mentions in fact two times, when the job is done : main screen and detail screen and both times have a small difference, which is a bit ridiculous.
    If I would see this at work in my applications, I would call our computer department to correct this, because this difference causes all kinds of unnecessary questions or reactions of users :
    - some users will laugh at it, because the computer can't count.
    - some users will think, which one of both times is correct and ask for it?
    - some users will think, if the time is wrong, what else is wrong?
    I know, it's hairsplitting, but I work with users all the time and our computer department is quite isolated from the real world. So they don't hear user's reactions.

    The button "Hide Details" is an unnecessary gadget IMO, just show the whole screen.
    The box "Close when task has completed" is always marked and until now I couldn't find a setting to change the default value of this box.
    I like to see the results when the job is done and I always have to unmark this box, if I don't forget it.
     
  9. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    HI Erik

    I just checked. I still didn't see the time on the main screen, I'll have to look again, unless you are talking about the upper and lower half of the copy box. In that case the upper box time only changes at increments. What I did notice interestingly is the final time in the detail box was 3 seconds different the what is in the log file. But that could be a function of the log file closing.

    I totally agree with you about the Hide Details.

    Another significant test for you would be to compare the time to either make an Acronis image or increment, and compare that to the time to refresh an FDISR archive.

    Cheers,

    Pete
     
  10. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    After the job is done the upper part mentions the time indeed, something like this : 13 minutes 6 seconds.
    The time is also mentioned in the lower part and they are often different, maybe all the time :)

    I'm not familiar with refresh yet, an explanation would speed up my testing.
     
  11. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    The heart and soul of FDISR. Okay, you have your primary snapshot, and you build secondary, by doing a copy and selecting secondary as the new snapshot, or archive as a new archive.

    To do a refresh just do a new copy, and select primary as source, and either secondary or archive as the target. FDISR will then do essentially a refresh,making the secondary or archive, look just like the primary. Note the target can't be the snapshot you are in.

    So say you want to uninstall your current security software to test say Norton's new suite. What I would do is refresh my secondary, and then have at making the changes. Then if I really like the changes and wanted to keep them, I might refresh both the secondary and archive. If I knew it was a mistake, I'd boot to secondary, and refresh primary. Then all changes are gone, like they never happened. Finally, if I wasn't sure and wanted to keep the changes temporarily, I'd boot to secondary, and create a new separate archive from Primary. Then refresh primary from secondary.

    There are almost as many different ways to use this beast as you have imagaination.

    Pete

    PS if unclear fire back at me.
     
  12. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    Peter,
    So if I install Online Armor in the secondary and I like it.
    My source will be the secondary and my target will be the primary and when the primary is refreshed the primary will contain Online Armor as well.
    Is that how it works basically ? If yes, I got it.
     
  13. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    Exactly. And if you went the other way it would be as though Online Armor had never been on your computer.
     
  14. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    OK. I got it, it saves me alot of reading and translating LOL.
     
  15. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    Hi Erik

    I figured out what is going on with the time stuff. The top window is a clock time, and agree's with what is written in the log file.

    The time in the bottom windows is an estimated/calculated time based on data transfer rate and data to be moved. It can't be clock time because it is estimating time to completion. Ergo they will always be off a bit.

    Pete
     
  16. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    Yes, but that's no excuse for me for not correcting the program.
    Each processing program has a start-time (= computer clock time) and a stop-time (= computer clock time).
    The "REAL elapsed time" = stop-time minus start-time.

    Once the program is finished, the only thing the programmer has to do is replacing the "ESTIMATED elapsed time" with the "REAL elapsed time" on the screen and in that case both elapsed times are equal on the screen.
    The user doesn't need to know this, he can't see what happens behind the screen, he isn't fooled either, because he has the "REAL elapsed time" on his screen and he has no reason anymore to worry about the difference.
    We use that little trick in our computer department too, a programmer can put ANYTHING on a screen, no matter what the program does.
     
  17. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    Erik

    To tell you the truth I've never paid attention to that. There is only one number on that whole screen that is really worth paying close attention to. Going to let you guess what that is.

    Pete
     
  18. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    I can understand that, but I'm trained to pay attention to these things and believe it or not some users pay attention to these things too, especially when they have to work with an application all day long.

    Besides, I like to know how long a processing program runs.

    If I would work with FDISR and I do the same with ATI, I would pay ALOT OF attention to the "Source object" and the "Target object".
    In case of FDISR the "Source Snapshot Number" and the "Target Snapshot Number",
     
  19. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    I bet not many users have even noticed it. The 2 or 3 second difference just
    never seemed a big deal.

    You for sure want to be sure your source and target are right, but once you have the copy detail window up, really the key number to watch is the error number. It should be 0, and if it isn't for some reason, the smart thing is to rerun the copy, or understand what is going on.
     
  20. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    For the first time, I restored an image with FDISR and two snapshots (primary and secondary) on it, over an existing partition with FDISR and six snapshots on it, without any problems.
    ATI restored the .tib-file with a required reboot and when ATI was finished, my computer rebooted and the FDISR boot screen (+F1-key) showed up as normal and Windows started normally.
    Another successfull test and a pratical proof that ATI and FDISR work together.
    Unbelievable, but true and a fact is more respectable than the lord mayor of London, as my English teacher always said. :)
     
  21. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    For the very first time, I was in trouble with ATI/FDISR. Keep in mind that I'm still trying to break ATI in every possible way and I only know FDISR for two days.

    I restored from a .tib-file without FDISR on it over an existing partition with FD-ISR on it.
    After the restore my computer rebooted and then I saw the FDISR Pre-boot Screen (+F1).
    How was that possible? And Windows was working properly.

    I rebooted and the FDISR Preboot Screen (+F1) was still there.
    I pressed the F1 and got a red message

    $ISR folder not found.
    FirstDefense-ISR ERROR - Press any key to reboot Windows.

    So I hit ENTER, which belongs to the family Any Key and Windows started normally.

    I said by myself "It can't be that c:\$isr\0\$isrbin-thingy", because it was never there and the red message confirms it. Maybe it was the MBR-thingy."
    So I restored the MBR-record of that specific .tib-file, using Acronis True Image and that worked. Everything back to normal. Big sigh of relief.
     
  22. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    Peter,
    I don't know if it is the same as Error=0, but I activated something else in FDISR.

    I clicked on Tools/Options/Tasks
    and marked the option "Verify destination snapshot after copy"
    I clicked on Tools/Options/Schedules
    and marked the option "Verify destination snapshot after copy".

    And this verification really runs, although very fast. Much faster as in ATI.
    But I guess checking "Error=0" won't hurt my feelings either. :)
     
  23. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    Funny, but I've never used that option in the couple of years I've been using FDISR. Never had a problem as long as the errors were 0. Acadia may say more, but he tried it for a while(may still be using it) and said occasionally it reported errors that that copy didn't. Not sure what to make of that. I've never as I said had an issue with copying snapshots. I do however periodically take one of my backup archives and test them by restoring them to a new snapshot, and boot to them to check them. So far so good.

    Will be very curious at the end to see what you think. I've briefly tested Shadowuser, and a more thorough test of Rollback Rx, but have come home to FDISR.

    Pete
     
  24. ErikAlbert

    ErikAlbert Registered Member

    Joined:
    Jun 16, 2005
    Posts:
    9,455
    OK. Peter, I hope you have read my previous posts above the error=0 issue about my problems with the MBR-record, using ATI and FDISR together.
    I think it's the only case, where you can have a problem with the MBR.
    It was nothing serious, because my system still worked during that problem.
    I told you about it, just in case you would do a similar action.
    An average FDISR-user won't have that problem. :)
     
  25. Peter2150

    Peter2150 Global Moderator

    Joined:
    Sep 20, 2003
    Posts:
    17,039
    Actually I did miss it. You must have posted it when I was typing. I absolutely love your technical description "thingy":D

    Since you are light years ahead of me on ATI, I assume on restore you can control whether it replaces the MBR or not. Not replacing the MBR would certainly explain what you saw.

    This is why I keep things simple and try to avoid possible problems. I always disable preboot in FDISR when imaging. Also I only ever image the whole disk, and mentally zero out all the exotic options in ATI,IFW and Ghost 2003.

    But what you have found is good news. It means you can do a full disk image with FDISR on. This tells me my plan which is only image like every 6 months is viable, as once the image is restored, the snapshots can easily be refreshed from archives.

    Erik I may poke at you at times to keep you honest so to speak, but I have to say your experimenting is yielding very valuable information for members of the forum. Thanks!!

    Pete
     
Loading...
Thread Status:
Not open for further replies.