Restoring a clone to an appleRAID mirror

Ben Bass's Avatar

Ben Bass

06 Aug, 2012 07:56 PM

I am running into some difficulties with restoring a clone to an appleRAID mirror. I am currently setting up a new Mac mini server running Mountain Lion and want RAID the internal drives. I can clone to a single drive and boot just fine. However when cloning the same OS to an appleRAID partition, the computer does not boot. It gets stuck at the apple logo with the spinning gear, and will eventually hang, or fall back to another bootable OS. On a different system I was able to get around this problem by doing a Time Machine restore to the appleRAID volume. This allowed it to boot, but as that was an Snow Leopard Server, other problems came up. However I was able to get that system working by erasing the disk, not destroying the RAID and partition, and then cloning back the working OS.

Any ideas on how to clone an OS to an appleRAID partition and get the computer to boot? The odd thing is that this used to work with CCC back in the 3.3.7 days.

  1. Support Staff 2 Posted by Mike Bombich on 06 Aug, 2012 09:10 PM

    Mike Bombich's Avatar

    Hi Ben:

    Can you submit your logs to me for review so I can see what you've tried so far? The easiest way to do this is from within CCC:

    1. Choose "Report a problem" from the Help menu
    2. Click on the "Submit Logs" tab and review the information presented
    3. Click on the "Submit Logs" button
    4. Update this discussion to let us know that you've submitted your logs, and please note the submission ID at the bottom of the Submit Logs tab.

    Also, are you using Disk Utility to create the RAID array, or something else?

    Thanks,
    Mike

  2. 3 Posted by Ben Bass on 07 Aug, 2012 06:45 PM

    Ben Bass's Avatar

    Hi Mike.

    I just submitted the logs with log id of 11103. In this last test I created the RAID as a single member using Disk Utility. I chose single member as the other internal drive is the one with the OS, and I didn't have another drive to clone to in the interim and then clone back to the blank RAID.

    Thank you!

    - Ben Bass

    On Monday, August 6, 2012 at 5:10 PM, Mike Bombich wrote:

    > **** PLEASE REPLY ABOVE THIS LINE ****
    > **** Any part of your response that is below this line will be discarded ****

  3. 4 Posted by Ben Bass on 07 Aug, 2012 08:55 PM

    Ben Bass's Avatar

    I just updated the logs. That initial clone did work. However the second batch cloning to an existing appleRAID did NOT. I wound up booting from an external drive running the same OS as the raid (10.8 server). I unmounted the raid and ran disk utility to repair the raid. I then repaired permissions and was able to boot from the RAID. This entire time the RAID was set as the startup disk - even though it booted from an external. The computer now boots successfully from the RAID after the repairs.

    - Ben Bass

    On Tuesday, August 7, 2012 at 2:45 PM, Ben Bass wrote:

    > Hi Mike.
    >
    > I just submitted the logs with log id of 11103. In this last test I created the RAID as a single member using Disk Utility. I chose single member as the other internal drive is the one with the OS, and I didn't have another drive to clone to in the interim and then clone back to the blank RAID.
    >
    > Thank you!
    >
    > - Ben Bass
    >
    >
    > On Monday, August 6, 2012 at 5:10 PM, Mike Bombich wrote:
    >
    > > **** PLEASE REPLY ABOVE THIS LINE ****
    > > **** Any part of your response that is below this line will be discarded ****

  4. Support Staff 5 Posted by Mike Bombich on 08 Aug, 2012 03:30 PM

    Mike Bombich's Avatar

    Hi Ben:

    I think I have a smoking gun. I see this in the log:

    08/07 16:10:15 Cleaning up...
    08/07 16:10:15 Updating the Mac OS X "Dynamic Linker Shared Cache" on the destination volume...
    08/07 16:10:15 Summary statistics:
    ...

    and a crash report for CCC's privileged task helper tool that occurred at that moment. The "Updating the Mac OS X "Dynamic Linker Shared Cache" on the destination volume" phase should take about a minute, but that appears to have flopped. If this phase doesn't run, then I would expect that the destination is probably not going to boot. It's easy to fix (and you have fixed it), but it's a problem I should have caught. I see this twice in your log, so it wasn't a fluke. I'm going to try to reproduce this one here and address the issue for an update.

    Thanks,
    Mike

  5. 6 Posted by Ben Bass on 09 Aug, 2012 12:36 AM

    Ben Bass's Avatar

    Thank you for the update Mike. If you need me to test a new build I will have access to the machine in question next monday 8/12.

    Thank you!

    - Ben Bass

    On Wednesday, August 8, 2012 at 11:29 AM, Mike Bombich wrote:

    > **** PLEASE REPLY ABOVE THIS LINE ****
    > **** Any part of your response that is below this line will be discarded ****

  6. Support Staff 7 Posted by Mike Bombich on 09 Aug, 2012 02:45 AM

    Mike Bombich's Avatar

    Hi Ben:

    I wasn't able to reproduce this here, unfortunately. One thing that might help would be to choose "Enable debug output" before running the task again. Assuming that that last phase finishes too quickly again, resubmit your logs via the "Submit Logs" tab of CCC's Help window. I'd then be curious to see the text produced when you invoke this tool directly in Terminal:

    sudo update_dyld_shared_cache -root /Volumes/Server

    Thanks,
    Mike

  7. Mike Bombich closed this discussion on 20 Sep, 2012 05:00 AM.

Comments are currently closed for this discussion. You can start a new one.