Surface 3 Fails to Image
I’ve been trying to get a Surface 3 (not a pro) to image using Fog for a few days now and have run into problems after problems after problems trying to just capture the image…
Currently I’ve done the following:
0) Installed SVN version of Fog (the logo on the webpage shows 5473) on Ubuntu 14.04 following the steps at https://wiki.fogproject.org/wiki/index.php/Ubuntu_14.04 using steps for SVN here https://wiki.fogproject.org/wiki/index.php/SVN (i only installed the SVN version
a) Side note this version seems a bit buggy, i can’t list all hosts (i get a server 500 error), multi-casting didn’t seem to work (All of the computers would load partclone and imaging would never start) and I can’t easily add members to a group (server 500 error) - It could be because i imported hosts/database…
- Disabled secure boot on the Surface
- Changed DHCP boot option to snp.efi (per this forum post: https://forums.fogproject.org/topic/4982/unable-to-successfully-pxe-boot-to-a-fog-menu-on-a-microsoft-surface-pro-3/10 )
- Create an image (I’ve tried both “Single Disk - Resizable” and “Mulitple Partition - Single Disk (not resizable)”
- Created a new host with the network adapter MAC address
a) I’ve tried setting the Host Primary Disk to /dev/mmcblk0 - Per this forum: https://forums.fogproject.org/topic/3662/surface-pro-3-pxe/65?page=1 ( I get the error message “cannot find HDD on system” otherwise)
- Boot the surface with microsoft network adapter (with a USB hub and keyboard)
a) I have to enter the tftp boot server manually
When the image is set to “Single Disk - Resizable” i get the following error: No resizable partitions found.
When the image is set to "Multiple Partition - Single Disk (Not resizable) - I get the following saying that the task is complete:
<Fog Logo> verifying network interface configuration...done checking oerating system... Windows 8.1 Checking CPU Cores...4 Send method....nfs checking in...done mounting file system...done, preparing to send image file to server. checking mounted file system ...done using image: <image name> preparing backup location...done looking for hard disks...done re-reading partition tables...done useing hard disk: /dev/mmcblk0 The protective MBR's 0xEE partition is oversized! Auto-repairing. Saving partitionTables (GPT)...done Task Complete! Stopping fOG status Reporter...done <tries to restart>
Not sure if this is helpful but here is the dmesg run from the surface ( I ran this without the Host Primary Disk set ): 0_1448392632297_dmesg.txt
Any help or links to forum posts i may have over looked would be greatly appreciated!
I tried imaging and it works! Thanks for all of the work you guys do!
@wwarsin Any updates from your side? The inits have been improved a lot in the last month and so far we made most of the tablets out there work with FOG. Please upgrade to the latest and see if it works for you as well. Please let us know!
@Wayne-Workman no, but the streams have the same hard drive, the mmcblk0.
@drc0nc Oh, I thought you were working with a Surface 3?
usb to ethernet adapter with
ASIX AX88772Bchip usb2 to 10/100. (I’ve been testing out gig to usb 3 adapters but no luck booting so far)
Host Kernel Arguments
Host Primary Disk
I’m using the HP Streams. I believe you already updated the wiki pertaining to these. We’ve lost the ability for a little while but @Tom-Elliott got it squared away.
Wayne Workman last edited by Wayne Workman
@Sebastian @drc0nc Please let us know what kernel you used, what version of FOG, what boot file, what USB to Ethernet adapter (the model), what kernel parameters or HDD parameters, and anything else special that you’ve done to get this device working.
I’m going to add this device to the Wiki’s working hardware list, that’s why I need all that stuff.
@Sebastian-Roth Works just fine now! thanks guys…
@Sebastian-Roth the write cache problem was only a warning, but I did come up with a fix I think anyway. There was no error code returned that showed there was or wasn’t a problem, so the fix I’ve added is to check the variable we store the check in. If the value has write cache with a value of not supported or is blank we won’t try enabling write support to begin with.
@drc0nc Thanks for letting me know! @Tom-Elliott Seams like we need to look into this again and find a way to check for writeCache without it failing on mmcblk devices. Any ideas? Maybe check the return code of hdparm?
@Sebastian-Roth I’m using the streams on this, and they’re running into the exact same problem with the mmcblk0 drive.
after running both hdparm -i and hdparm -W i was getting failed invalid argument.
@wwarsin Thanks for reminding me on this! Can you please go for another debug session and run hdparm with a different parameter to see how this behaves:
hdparm -W /dev/mmcblk0
@wwarsin Interestingly we got fog 1.2.0 to image a Surface 3 the other day. I couldn’t understand why trunk wouldn’t work
Have you updated recently and tried again?
@Sebastian-Roth Any updates by chance?
Here’s the output of the command:
# hdparm -i /dev/mmcblk0 /dev/mmcblk0: HDIO_DRIVE_CMD(identify) failed: Invalid argument HDIO_Get_IDENTITY failed: Invalid argument #
@wwarsin Great! Can you please run another debug session and let us know what you get from
hdparm -i /dev/mmcblk0
Yeah, pretty close we are indeed! :-)
@Tom-Elliott I think center alignment of those pieces in that picture below don’t look very good. I suggest making only the initial logo/credits box be centered, and everything else left aligned.
So close! Now i’m getting the following error when i try to deploy:
Checking write caching status on HDD…Failed
Could not set caching status (enableWriteCache)
Edit: I still have “/dev/mmcblk0” set as the Host Primary Disk
Success! It’s capturing the image now!
I’ll write back a little later, i’m going to try a deploy after this finishes to verify it works completely.
I appreciate all of the help you and Tom have been!