Most recent stable build worth using or fix for this error?
-
First, what is the host OS that FOG is running under?
Was this system running under the 1.2.0 prior to you upgrading to the latest SVN?I think we need to divide your issue into two parts
- FOG not installing correctly
- The image not deploying correctly (which may be a symptom of #1)
-
Hi sorry for confusion.
Story follows as
-
My original box was 3036 - working stable for some time, developed an error complaining about mounting nfs and checking for .mntcheck files - tried the sugggested fixes of touching the files but no luck. So decided to take a fresh box and make another fog server.
-
installed xubuntu 14.04.3 as stated above as host os. Installed latest svn on friday. Install died at dhcp setup. Figured it was a broken build so wiped my machine back to fresh manually and updated.
-
Tried with an older build 3559 - This installs absolutely fine. Created an image , added a host and performed an upload all with no errors.
-
Try to deploy and get
Erasing current MBR/GPT Tables… EBR Signature for logical partition invalid: Read: 0x000, but should be 0xAA55. Error reading logical Partitions, list may be truncated.
Hope that makes a little more sense.
I could I suppose try going back to 3036 though this build didnt move fresh images from the dev folder to their correct folder. Whilst not a deal breaker was a bit annoying.
-
-
@bengillam Hi, I’ve that problem on a linux image since few month. I’ve to upload the image in debug mode and run that before imaging :
fsck.ext4 /dev/your_linux_partition -
@ch3i said:
@bengillam Hi, I’ve that problem on a linux image since few month. I’ve to upload the image in debug mode and run that before imaging :
fsck.ext4 /dev/your_linux_partitionhi,
thanks so I need to run whole of the upload debug operation here -> https://wiki.fogproject.org/wiki/index.php/Debug_Mode
if so at what point should I run your command above.Thanks
Ben
-
OK slightly embarrassingly may have spotted the issue, a stray 5800 hp which physically looks identical was in the pile of fresh machines i decided to image expecting it to be a 7800 would probably not help! Have now got an actual 7800 which I am trying to image.
Also moved to svn build 4103 after reading some good things.
Running the image operation as we speak, will see if this works and post back.
Thanks
Ben
-
@bengillam said:
OK slightly embarrassingly may have spotted the issue, a stray 5800 hp which physically looks identical was in the pile of fresh machines i decided to image expecting it to be a 7800 would probably not help! Have now got an actual 7800 which I am trying to image.
Also moved to svn build 4103 after reading some good things.
Running the image operation as we speak, will see if this works and post back.
Thanks
Ben
ok that didnt work same issue with the EBR signature - however the image now completes and DOES boot (since 4103 upgrade). Is it safe to ignore the error or is there something deeper causing it.
-
@bengillam said:
EBR Signature for logical partition invalid
See fractal13’s post here: https://forums.fogproject.org/topic/2848/ebr-signature-for-logical-partition-invalid/2
Probably same in your case, trying to deploy to an completely empty disk, hmm? -
@Sebastian-Roth said:
@bengillam said:
EBR Signature for logical partition invalid
See fractal13’s post here: https://forums.fogproject.org/topic/2848/ebr-signature-for-logical-partition-invalid/2
Probably same in your case, trying to deploy to an completely empty disk, hmm?Hi yes, in this case had 5 machines supplied to us refurb with blank disks which I need to image. With my latest test finishing and booting don’t I need to worry?
Can certainly test with a machine with non empty disk.
-
I’m fairly sure we do attempt initializing the disk, but the basis of the fogpartinfo (which I would still love to find a more suitable method to get partition names), is likely breaking things right now.
-
@bengillam Hi, it’s before use the command
fogto run the job. -
@Tom-Elliott said:
I’m fairly sure we do attempt initializing the disk, but the basis of the fogpartinfo (which I would still love to find a more suitable method to get partition names), is likely breaking things right now.
so im a little confused, so the error im getting is a bug? In which case 1) is it safe to ignore if my resultant imaged pcs are booting, and 2) is there a more stable build you could recommend? (other than the much older 1.2.0 version)
Thanks
Ben -
- From my point of view it’s not a bug. Save to ignore.
- There is no stable build newer than 1.2.0 at the moment. But Tom has reworked the scripts and you are more than welcome to upgrade to the latest trunk and test things.
