Deployed image boot loop - black screen & Press Ctrl Alt Delete



  • Hi Guys
    Am running RC1.5.0-RC-9 on Centos 7.
    I captured an image of a Dell 3020 factory image by way of a backup so I could restore it if required. Captured with no problem (Win7-Single disk -multiple partitions-all default settings).
    I tried to deploy the image back to the same computer but on reboot I get a black screen with “Press Control Alt Delete to boot” when I do that it reboots back to the same screen.
    I’d really appreciate any insight into this. Thank you.



  • Okay. You have a point about backing up the original disk.
    I’ll give that a bash and check if it works.


  • Developer

    @TomcatKZN Discussing this with the other developers I came to the conclusion that using “Single disk - Resizable” in your case might not be a good thing anyway. You want to have a clean backup of the original state on disk as far as I understand. From that point of view you should just use “Single disk - Multiple partitions” which does not touch the partition layout. The Resizable method does work in many cases but there is some magic involved which you don’t want to rely on for your backing up the original disk state I suppose.

    I am not saying that we won’t fix things! But from my point of view you just shouldn’t be using the Resizable Image Type for what you are trying to achieve.

    I’d suggest creating a new image definition in the web UI using the “Single disk - Multiple partitions” image type and upload from a fresh untouched system again.

    Edit: Moved to bug reports!



  • Understood. Thank you.


  • Developer

    @TomcatKZN We have a proposal of a fix ready but I want to discuss this with the other developers before we release the fix. The resizing code is crucial and shouldn’t be messed with. Will let you know as soon as possible.



  • @sebastian-roth said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    @TomcatKZN Ok thanks, I think Ich found what’s causing this. Just need a little more time to fix. Will get back to you soon.

    Awesome. Much appreciated.


  • Developer

    @TomcatKZN Ok thanks, I think Ich found what’s causing this. Just need a little more time to fix. Will get back to you soon.



  • 0_1510056960842_dell-EPMOutput.jpg

    I also ran Easus partition manager and posted the output.
    Seems the 1st partition is fat16. perhaps thats the problem.
    Wonder if I can change it to fat32 if it will still work.
    What do you think?



  • @Sebastian-Roth

    GDisk -PreImage:
    0_1510043914797_Dell-GDiskOutputPreImage.jpg

    FDisk-PreImage:
    0_1510043953337_Dell-FDiskOutputPreImage.jpg

    Does this help identify the problem with capturing this image?



  • @sebastian-roth said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    @TomcatKZN Thanks for the pictures. Although I am sure we have found the issue we still need to figure out what exactly is causing this. I need more information on where those pictures were taken. I’ve tried to replicate this in a test VM but was not successful yet. So I guess my assumptions are wrong.

    Please let me know where and when exactly those gdisk and fdisk output pictures were taken. I surely see this is a FOG debug task but was this on a fresh untouched client before capture, after capture or on another machine before / after deployment?

    What I am sure is that the partition start positions are moved at some point. And because of that Windows cannot boot. Though I am still not sure if the initial partition layout is crappy or if this is something FOG is causing.

    The Gdisk and FDisk output is from a freshly deployed Client which failed.
    The same client from which the image was taken.
    The picture from windows is before the capture and deploy.
    I will install an untouched disk and generate the same output from before the capture.


  • Developer

    @TomcatKZN Thanks for the pictures. Although I am sure we have found the issue we still need to figure out what exactly is causing this. I need more information on where those pictures were taken. I’ve tried to replicate this in a test VM but was not successful yet. So I guess my assumptions are wrong.

    Please let me know where and when exactly those gdisk and fdisk output pictures were taken. I surely see this is a FOG debug task but was this on a fresh untouched client before capture, after capture or on another machine before / after deployment?

    What I am sure is that the partition start positions are moved at some point. And because of that Windows cannot boot. Though I am still not sure if the initial partition layout is crappy or if this is something FOG is causing.



  • Disk partitions in Windows:
    0_1510004520865_dell-WinPartitions.jpg
    Output from GDisk -l:
    0_1510004314404_Dell-GDiskOutput.jpg

    Output from FDisk -l:
    0_1510004568439_Dell-FDiskOutput.jpg

    Any help greatly appreciated.



  • @sebastian-roth said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    @tomcatkzn said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    d1.fixed_size_partitions = :1:2:3
    d1.minimum.partitions = same as d1.partitions

    Ok so this has essentially been more or less like “Multiple partitions - Single Disk” because our logic of detecting resizable partitions wasn’t able to figure yours as such and marked them all to be fixed size. Then again I wonder why you are seeing the “Press Control Alt Delete to boot” message on boot after deploying the image to another machine. Sure the Windows version and OS ID set in FOG do match?

    Please schedule a debug deploy task for one of your client. Boot it up and start the deployment by running the fog command. After it finished run fdisk -l /dev/sda and post a picture of the output on screen.

    Yes. Op sys is win7 and image type is set to Win7
    I watched the end of the deploy task and it said “Sync Successful” for each partition
    I restored it back to the exact same machine and still get "Press Ctrl-Alt-Del to Restart
    immediately after bios boot.

    Am running a debug task now.


  • Developer

    @tomcatkzn said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    d1.fixed_size_partitions = :1:2:3
    d1.minimum.partitions = same as d1.partitions

    Ok so this has essentially been more or less like “Multiple partitions - Single Disk” because our logic of detecting resizable partitions wasn’t able to figure yours as such and marked them all to be fixed size. Then again I wonder why you are seeing the “Press Control Alt Delete to boot” message on boot after deploying the image to another machine. Sure the Windows version and OS ID set in FOG do match?

    Please schedule a debug deploy task for one of your client. Boot it up and start the deployment by running the fog command. After it finished run fdisk -l /dev/sda and post a picture of the output on screen.



  • @sebastian-roth said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    @tomcatkzn said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    … Win7-Single disk -multiple partitions-all default settings …

    From what we see in the picture this was not captured as “Single disk -multiple partitions” but as “Single disk - Resizable”. So please post the contents of d1.fixed_size_partitions, d1.minimum.partitions and d1.original.fstypes as well.

    We had some issues with type=de partitions a while ago but I think those should be all fixed in RC version. By the way, do you use working or dev-branch?
    Not sure about the type=27 yet. Possibly this is causing the issue. Why is the boot partition type set to be a “Hidden HTFS partition”? I know this is what you got from Dell…

    So I’d suggest to actually create a fresh image in the web UI, set it so “Multiple partitions - Single Disk” and try capturing again from an original machine that hasn’t been touched yet. I am fairly sure this will work when deploying to another machine.
    Thank you again for the help. Sorry for delay. Needed to Eat :]
    I last used FOG 0.32 in 2012 - worked great, but I became ill and job went to someone else. That person switched to WDS and left without documentation. I’m a fan of FOG but see I have a lot to catch up on. The work youve done is awesome.

    Yes - my bad memory: it is indeed Single Disk-Resizeable
    I installed from dev-branch to get the latest build, as I am happy to give feedback.
    d1.fixed_size_partitions = :1:2:3
    d1.minimum.partitions = same as d1.partitions
    d1.original.fstypes = NTFS
    Am in the process of recapturing the image. Will post progress shortly.


  • Developer

    @tomcatkzn said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    … Win7-Single disk -multiple partitions-all default settings …

    From what we see in the picture this was not captured as “Single disk -multiple partitions” but as “Single disk - Resizable”. So please post the contents of d1.fixed_size_partitions, d1.minimum.partitions and d1.original.fstypes as well.

    We had some issues with type=de partitions a while ago but I think those should be all fixed in RC version. By the way, do you use working or dev-branch?
    Not sure about the type=27 yet. Possibly this is causing the issue. Why is the boot partition type set to be a “Hidden HTFS partition”? I know this is what you got from Dell…

    So I’d suggest to actually create a fresh image in the web UI, set it so “Multiple partitions - Single Disk” and try capturing again from an original machine that hasn’t been touched yet. I am fairly sure this will work when deploying to another machine.



  • @sebastian-roth said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    @tomcatkzn said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    backup the factory image from a single Dell machine

    To me this kind of sounds like a weird partition layout issue. Can you please boot one of the original machines into a debug capture task, hit ENTER twice to get to the shell. Then run gdisk -l /dev/sda, take a picture of the screen and post here. As well take a look at the image information stored on the FOG server in /images/<IMAGENAME>. Post the contents of the text file called d1.partitions here as well.

    Thank you for the helpful diagnostic comments. Just waiting for another install to finish and I’ll do the “debug” task.
    Here are the two pictures of/images/DellMaster and d1.partitions
    1_1509988882962_Dell3020ImagesList2.PNG 0_1509988882962_Dell3020ImagesList.PNG


  • Developer

    @tomcatkzn said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    backup the factory image from a single Dell machine

    To me this kind of sounds like a weird partition layout issue. Can you please boot one of the original machines into a debug capture task, hit ENTER twice to get to the shell. Then run gdisk -l /dev/sda, take a picture of the screen and post here. As well take a look at the image information stored on the FOG server in /images/<IMAGENAME>. Post the contents of the text file called d1.partitions here as well.



  • @george1421 said in Deployed image boot loop - black screen & Press Ctrl Alt Delete:

    @tomcatkzn Ok that is what I wanted to be sure of, you were deploying back to the same computer as where the image came from.

    So to collect a bit more info on this:

    1. Did the 3020 come with Win7 on it??
    2. Is the 3020 in UEFI or legacy mode.
    3. Is it possible to get a clear picture taken with a mobile phone and post it here?

    I’d like to see this error screen. I can only think of where this error might be, is during OOBE. But you are just capturing and deploying to the same hardware, so you aren’t running sysprep here. It should "just work"
    Hi George, Thanks for the response:

    1. Yes
    2. Legacy mode
    3. Nothing else is displayed before the black screen. Just the black screen and the white writing top left corner directly after bios post. I will try deploy the image again shortly and post the picture.

    I agree I am stuck. In the past a capture and deploy back to the same machine “just worked”

    Perhaps I try the capture again in case it is corrupted somehow.


  • Moderator

    @tomcatkzn Ok that is what I wanted to be sure of, you were deploying back to the same computer as where the image came from.

    So to collect a bit more info on this:

    1. Did the 3020 come with Win7 on it??
    2. Is the 3020 in UEFI or legacy mode.
    3. Is it possible to get a clear picture taken with a mobile phone and post it here?

    I’d like to see this error screen. I can only think of where this error might be, is during OOBE. But you are just capturing and deploying to the same hardware, so you aren’t running sysprep here. It should “just work”


Log in to reply
 

393
Online

39.5k
Users

11.1k
Topics

105.6k
Posts

Looks like your connection to FOG Project was lost, please wait while we try to reconnect.