Can Acronis clone Linux?

Hi - glad i found this site. An amazing amount of good info about Acronis - great help for a noob like me.

Here's a question I have regarding a need I have here at work: Can I, while in WinXP Pro, connect two Linux (LILO - occasionally GRUB) drives to unused SATA ports and clone from one that has our master image on it to a second, bare drive with the operation being run by Windows on the main system drive and have that cloned drive be bootable once it is cloned? Or does the file system on the source and target drives need to be in a NTFS/DOS/FAT format, which Linux is not of course.

We have an actual disk cloning machine made by Greystone Data Systems here in my lab but it - being pretty darn cutting edge - is VERY flakey in its operation. It can do a sector by sector "mirror" copy that doesn't care about the format of what's in the sectors. It doesn't make bad clones but the firmware on the darn thing gets hosed a lot and we're tired of constantly rebooting the thing. And it seems only stable at a DMA2 transfer rate. I'm thinking that we might install Acronis on a fairly beefy XP rig that I can get IT to build us and do it much faster - and with less grief! lol - if I can find a way to do this from a Windows desktop. Plus, then my engineers could clone Linux drives for development programs on their PC's from the comfort of their offices....saving me from constantly being awoken from naps when they barge into the lab all the time! Is this possible with Acronis?

Thanks!
Rick

 

No.

The last time I looked (and was able to find useful information on the U.S. site) there is Acronis True Image for Windows and Acronis True Image for Linux. If you want to work with Linux much, you would (have) want(ed) to purchase ATI for Linux. (I say, "would have" because it looks like U.S. support and website have gone belly-up.)

One thing you could try with ATI for Windows would be to create a bootable recovery CD and use that CD to boot up a Linux box. The recovery CD is Linux-based and knows all about *nix partitions. You should be able to clone drives with that, but there may be special drivers (and command-line boot options) required, and it will take a while because there will be no special chipset drivers (unless you load them at boot time).

 

I'm not wanting to work within Linux, per se. What I want to do is use Acronis on my XP system disk (call it disk 1) plug in two more drives (call them disk 2 and 3 so we know which drives I'm talking about) - one of which, the 'source" contains a Linux image - and tell Acronis that is running on disk 1 to use disk 2 (has a bootable Linux image) as the source of the clone and to use disk 3 as the destination/target. All of the examples I've seen online show folks cloning their system disk onto another drive. You pick which drive is the source and which is the target so it seemed to me that, if there were 3 drives, nothing said that the source disk HAD to be the one running Acronis.

I was thinking, if the "clone" process is a sector by sector exact copy, that it might not matter what format the data on the source and destination drives was in and that Acronis would just copy whatever is there to the same place on the destination drive. The Linux drives are used in one of our products and won't be booted in the XP PC that makes the clones. Or is it that Acronis has to use whatever volume/drive/partition it lives on as the source and you can't pick a different drive/volume as the source?

Rick

 

The proof is in the pudding. Why don't you just put those drives in there and give it a whirl? If if works, great!

Post back with your results!


P.S. ... the reason I remain skeptical is because Acronis True Image for Windows is a Windows application. No matter how smart Acronis thinks it is, in the scenario you describe, it is Windows running on top of your BIOS and presenting whatever hardware it "sees" to the operating system. I am skeptical that Windows is going to be honest about what you have mounted in those drive bays.

 

Hmmm, that's what I bought ATI to do about 10 versions ago (copy/clone Linux drives)--I think it should do what you want, unless you're using some filesystem version not supported (check ATI's specs).

 

i am also interested..
i would like to know if it's possible to clone a mac formatted partition from XP in a dual boot system...

lets keep the thread alive with news....
matteo

 

I just did several quick cloning tests in a VM and it worked fine.

Test setup:
Virtual PC
Ubuntu 7.10
Disk setup as Ext3 partition and Swap partition
GRUB installed to MBR.
Used TI 11 (8,101) booted into Vista Ultimate SP1.

Cloned to identical sized drive, partitions not changed. Booted okay.
Cloned and resized both partitions. Booted okay.
Note that I did not move the start of the Ext3 partition (the first partition).

 

Thanks, all! I haven't tried it yet because i'm waiting for IT to bring me a "guinea pig" pc to experiment with, but your comments and some reading of the manual make me confident i can pull this off. The manual says that ATI supports the Linux file structures, "Ext2" and "Ext3" so that made me confident as well. Our customer's Linux image is based on RedHat 9.1, so Ext3 should cover that. Actually, Ext3 is backward compatible with the Ext2 format so it's all good. Ahhh....sure would look good on the ol' review next Spring - might even be able to afford more than half a tank of gas if my boss is REAL happy! lol

I'll let you guys know how it turned out next week when I go back to work. Mon or Tue I hope - c'mon you IT dudes, hurry up and bring me the "patient'!

 

what about hsf? (mac format= would it work in the same way...?

 

Hey MudCrab,

Ok, I used one of our military PC's here but was unsuccessful in cloning at least on the first try. In this situation, though, I was cloning from a 40Gb IDE drive with Linux on it to a 160Gb SATA drive. I didn't mess with the new proportions as to the sizes of the 3 Linux partitions. It changed the size of the boot partition and the data partition but left the "swap" partition the same.

I'm trying again after setting the boot partition to the same size as the original disk's boot partition. TI says it won't be bootable without a Linux boot diskette (not an option) but I told it to go ahead anyhow. We'll see what happens. When I set the boot partition size on the destination drive, it ended up with about 360Mb on unallocated spce between the boot and swap partitions. Wasn't immediately obvious what to do to move this space (as I think, maybe I could have added it to the size of the data partition and that might have made it "allocated"(?)) so I left it alone.

I'll try a couple more things and then try cloning between same-sized disks, albeit one will be IDE and one will be SATA...

Rick

 

Well, nogo as far as cloning a bootable copy of a Linux drive even when using the exact same size/make disks. At boot, I get half a page of "99" followed by ye olde "Operating System not found". Dang! I'm going to borrow a usb floppy with a mbr zeroing program on it to see if maybe I need to wipe the master boot record of the Destination drive on top of doing a "Delete Partition" to it in XP.

Tried Automatic and Manual cloning but no joy.....any ideas?

Rick

 

Rick,

I got the same "warning" message from TI when I did the clone. However, it still booted properly.

What boot manager are you using (GRUB, LILO, etc.)?

Is it installed to the MBR?

 

The image is Lilo - that's about as far as my Linux knowledge goes. i'm fixing to try to zero the master boot record of the destination drave after deleting its partition to see if that is the problem.

One other possible monkey wrench in the works:The PC I'm using has Windows running on a SATA drive (SATA 1 in BIOS - ICH mode 'Enhanced-SATA').

The source, Linux, drive is an identical 160GB SATA laptop drive that lives on a board that interfaces the drive's SATA to a PATA connector on our motherboard. It shows up in BIOS as Primary IDE Master.

The destination is, again, an identical sized.make disk that is in an external e-sata/ubs enclosure (using the SATA not the USB interface) that plugs into our second SATA connector - "SATA 2".

I wonder if the difference in IDE/SATA interfaces is causing a problem? Gonna have to keep digging...

Rick

 

A lot of laptops use non-standard disk geometry. In those cases, the solution is to make sure the destination drive is installed in the laptop when the clone is performed. Cloning it outside the laptop won't work.

 

No, the PC isn't a laptop - merely used that term to describe that it is one of the small type drives that normally would go into a laptop - 2.5", I think.

The whole PC is non-standard, especially the motherboard is our own design. Intel based. I beginning to think that the cloning operation isn't cloning the mbr or boot partition exactly as I've read that Lilo is much more "picky" than grub about that stuff. Or maybe going across formats from and IDE interface to a SATA interface is a problem....not looking good. I'm going to try a Windows clone from the same ports with the same drives and see if THAT works...

Thanks for keeping tabs on my flailings here, btw...
Rick

 

Here's another piece of data - After cloning, I boot into WinXP Pro with the two Linux drives removed from the boot list still and open Disk Management and note that, although the two Linux drives have the same number and sizes of partitions, the good (source), bootable Linux drive has its first of 3 partitions sort of lightly greyed out with diagonal lines. The cloned drive doesn't have the greyed out bit. On both drives, XP reports these first partitions as "active" and the other two partitions on each drive as 'unknown'. The unknown part makes sense but I'm wondering what the greying out of the first partition on the drive that works means (probably why it boots and the clone doesn't) and why the clone's drive doesn't have that attribute.

BTW - I've tried Auto and Manual cloning, zeroing the MBR and "As Is" and "Proportional" partitioning with equal lack of success. I also tried a brand new, never been out of the bag drive. I'm using ATI 10 and wondering if I should maybe give 11 a shot if it has more capabilities and might work better for Linux cloning(?)

Many times, I will just be "restoring" an archived .tib image to a bare or wiped/part. deleted drive - which I did at home to my RAid 0 array and "assume" will work - but I need also to have the ability to clone Linux drives that have a lot of tweaks from our development programs onto new drives so that more than one engineer has a copy to "play" with, so that's why I'm gnawing so hard on this bone....

Thanks,
Rick

Rick

 

Okay, but realize that the problem could still be the same. You may not be able to clone in another computer and have the drive boot. If it's possible to do a clone (or reverse clone) in the original computer, that would tell you if it works or not and then you could proceed to test cloning the drives using a different computer.

I've only tested GRUB and it worked just fine. I have never used LILO, but I've heard it can be picky too.

I don't think that should cause a problem with actually booting the drive.

The gray diagonal lines usually just highlight the currently selected partition. You can verify this by clicking on another partition and seeing if the lines move.

I used TI 11 in my tests, but it was GURB too and not LILO. For imaging and using GRUB, TI 11's ability to restore in Sector-by-Sector mode might help. I've used that with success when previously the GRUB code was wiped during the restore.

 

LOL - yeah, I figured out the grey lines bit. Felt like a dork for posting it here.

Actually, the computer I'm cloning in IS the one that the Linux image ships in. We also have XP images that we use in the lab for development of new features...because only the contract guy who comes in whenever he gets up (lol) knows Linux.

I've been able to make an Acro full backup .tib of the Linux image and "restore" it onto a drive using the "rescue cd" method but I've come close to giving up on doing a "clone" of this drive image. I've tried just about every variation possible. Most times TI says that it is dorking with the boot partition and I can't figure out how the heck that can be because the drives are the exact make/model and size and I don't change anything at all as to the partitions' sizes. It must be that LILO is just too sensitive somehow. The military usually lags far behind the technology curve because they need to make VERY sure that that tech is "mature" and bug-free before they risk lives with it. I reckon that's why it's not GRUB.

The thing is, we have an actual cloner made by Greystone that will clone this image - both "smart copy" (only the data and not the empty space) and "mirror copy" - just fine, albeit slowly. ATI must not be doing an exact, sector by sector clone. We were intrigued by how much faster we can clone a drive in a fast PC than the Greystone and hoped to come up with an alternative for use here in the engineering lab using ATI. I'm still going to use it on a dedicated PC (if I.T. would ever get around to building it!) with a library of full backup archive .tib Linux images on a spare drive as that would cover the great majority of instances. If I have to, I can first make a .tib of a unique Linux image and then go the "Restore" route. Just takes a good bit longer and I'm wanting to really blow the doors off the Greystone time-wise. Windows clones and restores just fine in every configuration I've tried.

Well, stick the Linux on the back burner for awhile....next up is to try to clone to/from a solid-state drive....of a different size, no less. Should prove "interesting" because I tried that today and it didn't work but I think the guy's SSHD was faulty - a lab at the other end of the building couldn't do it, either but I don't know what they tried to use to do it. I don't think the guy who brought it in ever actually booted it before i tried to clone a new image onto it. I'll try to do a "restore" to it....

Thanks, Mud - I'll probably be adding more to this thread after I think about Linux for awhile. I'm going to try ATI 11 tomorrow - maybe it has changes that will enable LILO Linux cloning.....ever the optimist!
c-ya!
Rick

 

Ok - Don't know if anyone is still monitoring this thread but, just to disseminate the data, I was finally successful in both doing a "Restore" (with "rescue CD") of a Linux drive .tib image AND also was able to get a bootable copy thru the "Clone" function.

The problem was mostly that our bootloader was LILO, which is pretty picky about its surroundings compared to GRUB. Whatever slight imperfection in the clone/restore/partition process or maybe slight differences even in same make/model HDD's seems to freak LILO out.

What I had to do was make a Linux boot floppy with our PC running Linux that had our specific kernel version. You have to do a "mkbootdisk --device /dev/sda "your kernel version"" and then boot to the floppy with your clone below it in the boot order and type, "vmlinuz" at the boot prompt. This boots you into your Linux OS. From a root terminal/console you then just type, "/sbin/lilo", and it loads your kernel and a couple of other things. You exit out of the terminal and shut down. After that, the drive is bootable and works fine. Cool beanz!

(BTW - a caveat that is maybe just specific to our customer's drive image: If, when you're trying to boot the drive with a Linux boot/emergency floppy, you don't install the cloned/restored drive into the same ide/sata southbridge port where either the source drive was when you cloned it or where the source was installed when an image file of it was made, the drive MAY fail to boot fully into your Linux OS and hang with a kernel panic.)

The clone/restores were done with both same-size/make/type drives and also
I've even been able to do a "Restore" of a .tib full image of a 160GB rotational HDD onto a ~60GB SSHD (solid-state, or "flash", hard drive) made by super Talent. (haven't tried "Clone" on that one, yet) I had no problems with WinXP clones/restores although, when I cloned from the small drive to the large drive, it took a LOT longer to finish. By a factor of like 4. Cloning from big to small was as fast as all the others. Like 5 minutes for a 160GB image.

I had to make sure especially on the dissimilar sized SSHD not to resize the Boot and Swap partitions. All of the restores and clones were done done "Manual" with manual partitioning. This was since, with "Proportional", the size of the boot partition would have changed on the dissimilar sized clone targets or destinations.

So, I've pretty much been able to clone anything I've gotten my hands on, Windows, Linux, and even a two-disk RAID 0 array and get bootable copies.

I have maybe noticed one thing that I'll probably need to start a new thread on: When I made full backup image files, I just manually saved them onto the root, C drive of the WinXP disk that had Acronis on it. I didn't do any of the "Secure Zone" crap. This worked ok, it seems, UNTIL....I made a second backup image of a totally different drive with either a different version of Linux or with a different OS. When I was in Acronis booted from the safe version of the rescue disk and wanting to do a restore, I was prompted to browse to the archived backup file location. As soon as I clicked on one of the two (differently named, btw) .tib files, Acro crashed to a black screen with a blinking cursor/dash and never budged. I'm "guessing" that maybe ATI doesn't like to see more than one full backup .tib file on the same drive or maybe partition? I tried moving the .tib's to different folders off the root of the drive but nogo.....but that, as they (sorta) say, is another story/thread...

Hope this is of help to someone, somewhen...

Rick