You are not logged in.
Hello ,
on new comput er
have a triple boot system Windows 11 Enterprise - Ubuntu 24.04 - Devuan 5.0/6.0
all 3 system have a proper NVME ;
now had some issue with Win11 which prompted me to re-install Windows 11;
since those NVME are small and pesky, did not bother to remove them manually (at first) from motherboard.
encountered steps on the way, which baiscally prevented me from installing Windows 11 again on the same NVME;
could not properly figure out where the problem could lie; thought it would be because of using USB installer, but it was the same with DVD drive;
would not want to remove other NVME since that should not be necessary;
what seemed to help was to change setting in BIOS > Secure Boot > from "CUSTOM" > "OTHER" - back to "WINDOES UEFI" >> afterwards instaling Windows 11 went through without no proplem; but could not longer boot the other OS, so i had to change back to "CUSTOM" > "OTHER" and it seems that Windows is happy algon the way nevertheless;
now my problem;
in the hustle, several time deleting and installing again Windows on the same NVME , i must have accidentally once clicked too fast and deleted a partition from another NVME , either Ubbuntu or Devuan - one;
now the Linux, sti.ll boot in OS, however, with Devuan there are several issue, which i hope to clear with your help;
basically .
1. it takes very long to boot (this problem is indipenddent of error i laid out above)
there is delay on boot due to
like [0.805369] hub 10-0:1.0: config failed, hub doesnt have any ports (err - 19)
2. and boot after this delay stalls with following error:
fsck.fat 4.2 (2021-01-31)
fsck exited with status code 6 (File system check failed
so most probleably i might have deleted the 1st partition on the NVME for Devuan;
please you be so kind to show how to recover from this error;
thank you.
Offline
Boot your Devuan installer and reinstall the boot manager from the Expert menu.
Offline
thanks , have re-installed, EID: i meant re-installed Devuan 5.0 entirely, problem solved;
will remember your suggestion, could need another time;
tried ubuntu-boot-repair and rescatux, to no avail.
Last edited by kapqa (2025-05-29 17:54:42)
Offline
Each time you format a partition or drive, the partition gets a new UUID.
The old UUID's listed in grub may be different, takes longer to find, and check.
You may edit grub to boot, then fix the problem with update-grub on each OS.
I hope this helps next time.
pic from 1993, new guitar day.
Offline
Hi, hope you find a solution! Some advice for anyone finding this thread:
Open your EFI partition in terminal then run
dd if=/dev/zero of=zero.img
then run
rm zero.img
then take a backup of the partition using dd or any disk utility then zip up your partition backup and save it for efi partition deletion emergencies. For accidental deletion of contents, just mount the backup image and copy the files over from it.
If you need to recreate a lost EFI partition, you can set the label to be the same as the lost one very easily.
Offline
MLEvD, did you copy/paste the wrong commands? That dd command will replace whatever is on the partition with a file containing zeroes. That would reproduce the OP's problem instead of fixing it.
Offline
dd if=/dev/zero of=zero.img ; rm zero.img
would just zero-out the unused space on the partition. Probably so it compresses better with zip.
Offline
rbit is correct, EFI partitions are typically mostly empty and are at the start of the disk where there can be lots of non-zeroed old junk that can make your compressed images much larger.
The joy of using dd in a fat partition where you never need sudo
Offline
Thanks. I got confused because I always use dd on whole devices or partitions. Sending it to a file won't clobber other files.
Offline
Yr welcome, let me tell you about my day(s) - spent all of yesterday upgrading from debian 6.0 CD1 lxde/openbox to Bookworm/LMDE 6 with sysvinit still installed from debian 6.0, didn't backup along the way as it was more about seeing what happens if kmod isn't allowed to upgrade and kick the 2.6 kernel, anyway, after the final upgrade, unbootable on all fronts, all kernels, all inits, and bash init was borked by a short libtinfo.
The day before, had upgraded LMDE1 to LMDE6 with sysv. Took a backup at Debbie and a final one after Faye.
This morning, couldn't boot system 2, so decided to restore system1. Faye backup was borked and unfixable, luckily Debbie backup worked and was fixable (needed swap repair).
I blame usrmerge, or kmod, or 18 yo eee pc ssd, or my 90% of backups untested, and another 9% untested.
Last edited by MLEvD (2025-06-08 11:19:59)
Offline
There is a beautiful solution for backups: fsarchiver (in my opinion, the best backup tool). It can back up everything—EFI, /boot, even LUKS partitions—while saving only the content, thus preserving space. The resulting backup files are small.
Offline