Acronis does file table based imaging AND RAW sector-by-sector imaging of the entire disk. On a side note if you zero out unused sectors, then do sector-by-sector backup, it comes out to be the same size as as a standard "allocation style" backup! Of course this would kill an RBRX volume. In acronis, a zeroed sector is represented by a single byte. In plain English, I had a 320GB disk with 150gb data on it. A file based backup came to be about 110gb with standard compression. A sector-by-sector disk image comes to 320gb with no compression. A sector-by-sector disk image comes to 110gb with standard compression and a "wipe free space" operation prior to imaging. Zero'd sectors take essentially no space! Without a wipe free space op, but with compression, it was about 230gb. Because of the abandoned unused data left behind.
Keatah, the question was whether or not Acronis uses the file-table when doing sector-by-sector imaging. Do you know if it does? pv
While discussing ZERO prepping of unused disk space I said... Just recently, while testing Drive Cloner v6 BETA under W8.1 and its ability to image RBrx snapshots, I discovered that ALL SSD Optimization functions have been disabled by Rollback... not just Windows TRIMming of recently deleted data (which they say they really, really do but no one in the technical community can get their heads around what actually is done), but also the SSD Optimization function which uses TRIM to insure that all unused space on an SSD partition is cleaned up and ZEROed to assist the SSD in proper NAND Block Mgmt. Since there's never been an adequate explanation of how TRIM really works while using a Rollback protected SSD (HDS says... "Don't worry... we really do TRIM your SSD, just not in the way you think its done."), my guess is that the Optimization TRIMming function that's attempted under Windows 8.1 (just like the Block Mgmt. TRIMming function under W8.1) just never happens. It's not clear to me why HDS can't really explain how they TRIM your protected SSD. With a pretty basic technical explanation I'm sure folks would understand the scheme being used without giving away so-called "trade secrets." Kurtis... see if you can get someone to come up with a basic WHITE PAPER on SSD TRIMming under W7 and/or W8.1. I'm sure that would help a whole lot.
Froggie, just use Dummy File creator http://www.softpedia.com/get/File-managers/Dummy-File-Creator-Portable.shtml or FileFiller http://www.softpedia.com/get/System/File-Management/FileFiller.shtml (create files with random data) to fill a snapshot or just add video files to fill a snapshots free space (leave at least 2gb free unless you are looking for trouble). Then boot to another snapshot and remove that large snapshot and perform a "Rollback defrag". Then use CheckFlash http://mikelab.kiev.ua/index_en.php?page=PROGRAMS/chkflsh_en and scan/read the partition for zeros. Then report here your findings. If you see a large part of the disk filled with zeros RollbackRX supports trim; if not... you know the answer. Panagiotis
@TheRollbackFrog I found this, if you'd like a bit more let me know (You know me, I'm always available!) Question: Does RollBack Rx support SSD TRIM Official Answer: Yes, it does. RollBack Rx has to support TRIM due it its architecture. 2nd Question: How come this is different from what I see when using TRIM check and other utilities? Answer: You need to understand that RollBack Rx fundamentally changes the “view” and “perception” of disk usage. On a system without RollBack Rx, the disk sector allocation is “black and white” (used or unused – absolute), there are used sectors and there are unused sectors. What you see is exactly what it is (binary). On a system with RollBack Rx installed, the disk sector allocation is much more complex and “colorful” (then black or white only). These are what RollBack Rx calls used sectors and unused sectors. But the used sectors can belong to multiple/various snapshots. A sector can be used by snapshots 25, 26 and 27. The most intriguing part, is when you are in snapshot 20, that sector could appear unused to the system. What you see is not what it actually is. So continuing with the example above, when you run disk diagnostic tools (TRIM) or defragment tools in the snapshot 20, you will not be able to make a correct judgment for that sector (you see it as an unused sector but that is not the case). If it’s not seeing the sector usage correctly, how can it make a correct judgment based on the wrong information? This is why we don’t recommend users to run disk defragmenter tools or rely on SSD TRIM apps after installing RollBack Rx. It is not that these apps will damage RollBack Rx, but because they are not going to provide you with a clear and accurate picture of what is going on. With regards to defragmenter we recommend uninstalling RollBack Rx or at least updating the baseline once you have completed defragmenting. With SSD TRIM utilities that work from within Windows you will not get an accurate read on TRIM. (Source)
Somehow I missed answering that. All evidence says it does not. It treats each sector as a separate entity. It will compress a group of zero'd sectors to almost zero size. You can have 1 tiny file on the disk, and all the other sectors full of random data. This is going to give you a big image almost full size as the source disk.
Well, now I have a fairly definitive answer. Although Kurtis supplied the "party line" from HDS and its developers, this is what I did to gain what information I could. 1. Using PARTITION WIZARD's "Wipe Partition" feature, I backgrounded the entire test SSD partition with ALL ONES. The result, an unformatted (unrecognizable) partition filled with ALL ONES. 2. Checked this status using BOOTICE's "Sector View/Edit" function (under WinPE) which allows me to see any block on the disk. The entire disk had been written with ALL ONEs. 3. Using Drive Cloner v6 BETA (an advantage of being in their BETA test group), restored a 4-snapshot single partition image of a Rollback protected system then BOOTed into the RBrx protected system just fine as has been tested previously. This process restores a RAW image of only the Windows used + RBrx Snapshot area... no additional blocks in the partition past the last one used by either of the aforementioned functions. 3. Using the same process as in Step #2, except this time under the RBrx protected Windows system, found the ALL ONEs data pattern in the unused area of the SSD. 4. Added some files, created Snap #5. Added some more files and ran the W8.1 Optimization feature on the test partition (supposed to TRIM almost all unused space on the partition), created Snap #6. Added some more files and created Snap #7. Rollback to Snap #4, deleted Snap #5-7, ran the Rollback Defragmenter. ReBOOTed the system and at Rollbacks' sub-console, stopped and ran the Rollback defragmenter once again. When finished, BOOTed into the Rollback protected System partition. 5. Repeated Step #3 with the same results... no TRIMming of FREE SPACE had been performed. 6. Repeated Step #2 and found the FREE SPACE still filled with ALL ONE's. 7. Restored a non-Rollback enabled version of the same system and ran W8.1's Optimize feature on the System partition. 8. Using the same process as in Step #2 except this time under the non-RBrx Windows system, found the ALL ZEROs data pattern in the unused area of the SSD as expected (This SSD TRIMs to ZERO... some others TRIM to ONE). Panagiotis, as both of us know, there really is no difference between the Windows TRIM command issued as a result of on-demand file deletion, and the TRIM commands issued during an Volume Optimization process. The way Volume Optimization works is Windows pre-allocates large blocks of FREE SPACE then immediately follows that up with deleting and freeing up that pre-allocated space... this causes Windows to issue TRIM commands for the "deleted" space. It does this in large blocks to almost the entire free space of the partition. The operative function here is the TRIM issued following the deletion of the pre-allocated space is the exact same TRIM function issued following an on-demand file deletion and deallocation of the space in use. Based on the above, I do not believe ANY TRIMming of data is occuring through the Rollback protected System.
Ahhhh! Froggie if you post such results you will get a response from Kurtis that: A) you are biased towards HorizonDatasys.... or B) fill a ticket support.... or C) trim is supported from RollbackRX and you are daydreaming.... Now, if you have time to spare, do RollBackRx users another favour and run the following batch file (with no admin rights) from a RollbackRX snapshot and after it finishes reboot the system. filldisk.bat of David R. Tribble http://david.tribble.com/text/filldisk.html Code: @echo off rem ============================================================================ rem filldisk.bat rem rem Fills the current disk drive with many large data files. rem This has the effect of wiping out all the unallocated file blocks and all rem the file blocks from previously deleted files (which may contain sensitive rem personal information) on the current drive. rem rem This is not a guaranteed way to completely wipe out all traces of your old rem data files, but it's pretty good. rem rem usage: filldisk D rem rem The 'D' parameter specifies the disk drive to fill, and is required. rem rem Note that filling an active disk drive beyond 90% of its capacity will rem probably cause MS/Windows to complain, especially if it is the Windows rem system drive. rem rem This works on MS/DOS 2.0 or later. rem rem 1.0, 2003-01-17, David R. Tribble, <david@tribble.com>. rem This code is not copyrighted and is in the public domain. rem ============================================================================ rem ==================================== rem Run in a temporary directory on the specified (or current) drive if not (%1) == () goto setdisk echo Fill the current disk drive with many large data files. echo. echo usage: filldisk D echo. echo The 'D' parameter specifies the disk drive to fill, and is required. goto end :setdisk set i=%1 shift echo Changing to drive: %i%: %i%: if exist fstemp\. goto skipmd echo Creating temporary subdirectory "fstemp" mkdir fstemp :skipmd chdir fstemp rem ==================================== rem Create a large file (128 MB) containing random data echo. echo Creating temporary data file: fs0.dat echo Booger> fs0.tmp echo Creating temporary data file: fs1.dat set i=a :loop1 copy /b fs0.tmp+fs0.tmp+fs0.tmp+fs0.tmp fs1.tmp >nul copy /b fs1.tmp+fs1.tmp+fs1.tmp+fs1.tmp fs0.tmp >nul set i=%i%b if not (%i%) == (abbbbbb) goto loop1 :endloop1 rem ==================================== rem Copy the random data file to multiple files until the disk fills up echo. echo Filling the disk with random data files... echo (This should eventually fail when the disk is almost full) echo (If it does not fill the disk, run it again in a different directory) rem These nested loops will create 10x10x10x10 = 10,000 files rem This gives a total of 128MB x 10,000 = 1,280,000 MB = 1,250 GB set i=0 :loop2i set j=0 :loop2j echo. set k=0 :loop2k set m=0 :loop2m echo writing: fs%i%%j%%k%%m%.tmp copy /b fs0.tmp fs%i%%j%%k%%m%.tmp >nul if (%m%) == (9) goto endloop2m if (%m%) == (8) set m=9 if (%m%) == (7) set m=8 if (%m%) == (6) set m=7 if (%m%) == (5) set m=6 if (%m%) == (4) set m=5 if (%m%) == (3) set m=4 if (%m%) == (2) set m=3 if (%m%) == (1) set m=2 if (%m%) == (0) set m=1 goto loop2m :endloop2m if (%k%) == (9) goto endloop2k if (%k%) == (8) set k=9 if (%k%) == (7) set k=8 if (%k%) == (6) set k=7 if (%k%) == (5) set k=6 if (%k%) == (4) set k=5 if (%k%) == (3) set k=4 if (%k%) == (2) set k=3 if (%k%) == (1) set k=2 if (%k%) == (0) set k=1 goto loop2k :endloop2k if (%j%) == (9) goto endloop2j if (%j%) == (8) set j=9 if (%j%) == (7) set j=8 if (%j%) == (6) set j=7 if (%j%) == (5) set j=6 if (%j%) == (4) set j=5 if (%j%) == (3) set j=4 if (%j%) == (2) set j=3 if (%j%) == (1) set j=2 if (%j%) == (0) set j=1 goto loop2j :endloop2j if (%i%) == (9) goto endloop2i if (%i%) == (8) set i=9 if (%i%) == (7) set i=8 if (%i%) == (6) set i=7 if (%i%) == (5) set i=6 if (%i%) == (4) set i=5 if (%i%) == (3) set i=4 if (%i%) == (2) set i=3 if (%i%) == (1) set i=2 if (%i%) == (0) set i=1 goto loop2i :endloop2i :end and report back how RollbackRX behaves in that situation. (I am curious if they have finally fixed with 10.x, the filling of the disk problem, that locks RBRX pre boot environment; you cannot boot to a snapshot, cannot change snapshots and cannot delete any of them. I reported it to them back in 2007 and until 9.x wasn't fixed.) Panagiotis
Hi I decided to try RollbackRX home edition but had a problem with the installation . After install the Pc boots to a safe mode type environment with faulty screen size over and over again , even after several un/reinstallations . The support FAQ e.t.c at HDS site gave nothing so i thought i should try here . Anyone knows whats up and how to fix it ? Oh , and another question - am i save from the potential problems using Rollback discussed here in the "special thread" if i also have a backup solution like Macrium on another partition that does full backups of C: .
If you image your system prior to the Rollback installation, you should be good to go. If you image your system after the installation of Rollback, you must do a COLD image (via the Macrium Recovery Media only) and use the "Make an exact copy of the partition(s)" option in the Advanced Options section of the imaging process. This will, of course, produce a larger image then the normal "used sector" type... be prepared.
pb1... sounds like a very unique problem to me. Your best approach would be to submit a SUPPORT TICKET at HDS... you'll definitely get more attention over there.
Can you take a screenshot or a photo of the monitor and post it here? Usually it happens when the RBRX subsystem installs in a fragmented state. Remedies defrag (the last protected partition) before installing RBRX and if it still gives the same problem you'll need to zero out the last 3GBs of the last protected partition.
No photo but i can paint you a picture . Think of your Os looking like safe mode , graphics wise , and , the picture(s) of it being 25% larger than your screen and thoose 25%`s vanishes on the right side of the screen . So the left side is ok but upside and downwise and to the right the screen is to short for the picture . I guess i will wait until , and if , Kurtis arrives here , or create a ticket after the weekend , or just skip the whole thing since i have had bad experiences with RBRX before and it still seems to be a dangerous/unstable software . The latest try was c:a a month ago and after extensive testing of all the programs features including uninstallation it freezed and gave me a bluescreen on the third uninstall .
Got it. I is different from the situation that I described above (that comes also with strange artefacts and corrupted buttons in the interface). You situation should be fixed if you select on your monitor the auto adjust key. It seems that RBRX uses a lower resolution that you never tried on that monitor and you need to calibrate it manually or automatically. Panagiotis
It appears to have been dealt with in some form. After I fill the disk and reBOOT into the sub-Console, it does fail to create a BOOT time snapshot (configured in RBrx) due to insufficient space, but does allow me to rollback to a previous snapshot which, of course, makes all that used space show up as available as expected. See... there has been some progress since 2007...
Kurtis here. Don't skip it please, we can get this fixed. Best bet would be if you submit a ticket and attach log files. Mainly because our tech support guys are much more technically inclined then I am and will take a look at everything for you. Do me a favour, and message me your Ticket ID. This is so I can audit the ticket and ensure that everything is dealt with properly and promptly. If you're in need of a response ASAP, include your phone number in your support ticket indicating that you'd like a phone call. That way, our tech support guys will call you. In some ways the phone call is easier. Cheers, PS: For future reference, don't hesitate to tag me. It'll help me get you a faster response.
Hi Thank you for your reply and offer . It was a flawed graphics driver that i fixed and now i can test the program perfectly .