You are not logged in.
I just built the package. It's in the ceres repo and will propagate to mirrors in the next few hours. Then it will move into chimaera in a week, I think. Note that this is only the desktop window theme and doesn't even include the desktop background image.
Note2: There's an earlier version I built a month ago. If you already installed that, it'll get upgraded to the new version when the new one moves down. Ignore the big jump in version numbers. That has nothing to do with the extent of the changes between the versions.
It's here if you absolutely can't wait.
https://pkgmaster.devuan.org/devuan/poo … -1_all.deb
Thanks. That's exactly what I wanted to know.
Yeah, every different login manager has its own config file where you can set stuff like that. You did it the right way. I'm not sure why it wasn't working right. One of the live-config scripts sets up autologin when you boot the live media, but it should not have done anything if you had autologin already set.
I just booted a live iso in qemu and installed vbox. I had to say those words in my head before I realized how weird it is. Anyway, I got the same error messages as you. Before I did it, I checked in sysv-rc-conf to see where dbus starts, and it starts in 2-5, not S. I checked beowulf, ascii, chimaera, jessie and wheezy. All the same.
There's no mention of dbus in /etc/init.d/vbox* or /var/lib/dpkg/info/virualbox-6.1.* so I don't know where that message is coming from.
I don't have any ideas for the usb. Yes I do. Not a fix; just a workaround. Someone I know used shared folders with a usb mounted on the host system.
I still want to know what you did to set up autologin so that I fully understand the cause of the problem, in case someone else does the same.
I have not read any accounts of anyone migrating from devuan to debian or other debian-based systemd distro. You would be cutting a new trail.
Obviously, you would need to add a mint or debian repo to your sources to install banned packages. You might need to do some creative pinning to prevent a disaster. Either that or do a full migration back to LMDE. Please write it up if you do. Thanks. Good luck.
Thanks for playing until the end. I like my stuff to work, and I'm really glad you stuck around until we all solved it. I might have to adjust the script so that mksquashfs is noisier about failing.
When you're in a mood to read some documentation, take a look here:
https://refracta.org/documents.html
Cool! Please tell us a little about what desktop environment and display (login) manager is installed. And if you have autologin set for your installed system, please describe how you did that. The live-config scripts set up autologin for the live environment. Maybe two settings are competing.
Optional: At the isolinux boot menu, press TAB to edit the boot line and add the one word, noautologin. And see if it boots to a login screen. Or maybe it'll boot to desktop.
^ definitely looks like a size issue, i wonder how large the snapshot would be if enough space were available?
I'm not sure how big it will be. It's possible to make the iso smaller by uncommenting one of the mksq_opt lines in the config file.
# Uncomment one of the lines below to use xz compression for smaller iso.
# small and slow
#mksq_opt="-comp xz"
# smaller and slower:
#mksq_opt="-comp xz -Xbcj x86"
One thing that might help is to edit /etc/refractasnapshot.conf to set save_work="yes". Normally, the copy of the filesystem gets deleted after the snapshot.iso is made. If the work dir is not deleted, we can do some forensic inspection and some other tricks, maybe.
Also...
I did not see any indication of a common problem in the log file you sent me, but looking again at your df output has me wondering if you're running out of space. You posted:
Filesystem Size Used Avail Use% Mounted on
/dev/sda1 8.6G 4.0G 4.2G 49% /
/dev/sda2 8.5G 3.0G 5.0G 38% /home
/dev/sda6 247G 98G 138G 42% /media/data
I see 4G of operating system plus 3G of home used. That's a total of 7G unless you added a bunch of stuff to the excludes file. I'm guessing you did not do that. Refractasnapshot makes a copy of the entire system (minus what's listed in the excludes file) and puts that copy in /home/snapshot. But you only have 5G free space in /home.
You can switch the location of work_dir and snapshot_dir in the config file. Change this:
snapshot_dir="/home/snapshot"
work_dir="/home/work"
efi_work="${work_dir}/efi-files"
To this:
snapshot_dir="/media/data/snapshot"
work_dir="/media/data/work"
efi_work="${work_dir}/efi-files"
Then run refractasnapshot.
This kind of presents the same hardware picture as your post #34 execpt that it has sda4 instead of sda6.
That difference is probably the cause of your problem... and now we need @fsmithred to draw conclusions
I don't see why that would matter. It should be booting the live system, not a disk partition. But it does lead me to a question. What's in the boot command?
Bob, at the boot menu for the live-usb, press TAB and see what it shows at the bottom of the screen. It should be
/live/vmlinuz initrd=/live/initrd.img boot=live username=computerbob
I can boot to initramfs prompt if I remove 'boot=live' from the boot command. I'm wondering if your boot menu got screwed up and lost that piece.
The prefix is hard-coded in grub-efi-amd64-signed. I'm not sure what changed in this last update. I first ran into this problem a year ago. If you remove the signed package and just keep grub-efi-amd64 it should fix the problem. Usually. And not if you need to use secure boot.
Assuming this is your wired interface:
dhclient enx00243218dfaa
Ok, the problem is in /etc/network/interfaces. Replace eth0 with enx00243218dfaa or do something to change that long name. Is that a usb ethernet dongle? You shouldn't see interface names like that for internal interfaces.
I had a similar problem, and Ralph's advice helped, but I had to delete those links a few times.
There are three hard drives in this computer. Up until a few months ago, when I plugged in a usb, it would be named /dev/sdd. Then one day it changed to /dev/sde. I've seen this in the past after pulling out a usb without unmounting it, but that always clears with a reboot. This has persisted.
I looked in /dev/disk/* and there were two links to /dev/sdd. One was 'Generic usb something' and the other was 'pci-something' that corresponded to the usb controller slot. I deleted those links two or three times, and they kept coming back. Then I deleted /dev/sdd along with them. That seems to have done it.
Thanks for persevering, dice. I gave up too soon. I did an install with targeted drivers and the initrd was only 8M, but maybe because the hardware was virtual. Let's see how this turns out, but I think you may have solved it.
Confirmed that ComputerBob's problem is not the same as yours. He has a full initrd.img (24M).
24M is big, and you confirmed that you didn't select for targeted drivers only. So that means your problem is not the same as dice's.
@ComputerBob:
When you return to this, here's a simple check to see if your problem is the same as dice's. Adjust the file name if you have a different kernel.
ls -lh /boot/initrd.img-4.19.0-14-amd64
Do you recall if you did an expert install and chose to only have drivers targeted for your hardware in the initramfs? If you don't recall, that's fine. The output of the command above will tell us. It'll either be big or small. (I want the number, please.)
I think I found it. Move /etc/initramfs-tools/conf.d/driver-policy out of the way and run update-initramfs -u and then make a new snapshot.
Unpack the initramfs from the bad iso and also the one from the good iso and compare them. (Hint: meld is a good program for comparing files and directories.) Use unmkinitramfs instead of the commands I posted in that other thread. It's easier. Plus, the command I posted for computerbob won't work for your initrd.img.
Yeah, the second way is correct. I should have explained the quotes and $. Replace that with path to the file you're checking. And it looks good. The script removed crypttab and resume as it should have. I don't know what the next step is.
I have a couple of chimaera-based live isos here:
https://get.refracta.org/files/testing/
No xorg. No non-free. Run 'refractainstaller' to install to hard drive. Or do a debootstrap install.
Please do the same on either the mounted iso or mounted usb stick that has the image. Check /live/initrd.img to see if resume and crypttab got removed. Is the host system that made this snapshot an encrypted system?
lsinitramfs "$initrd.img" | egrep "resume|cryptroot|crypttab|askpass"
Note: in the previous post I started the command with '#' which suggests it needs to be run as root, but that is not the case. You can run the command as user.
Sorry for the tease. That link might be one of my special powers as admin.
I don't understand the symlinks at all. When you looked at the mounted usb with ls, it showed the right things.
You can mount the iso as root and look at it that way.
mount /home/snapshot/snapshot.iso /mnt
ls /mnt
ls /mnt/live
ls /mnt/isolinux
# and when you're done looking around in there
umount /mnt
You should see /live /isolinux /pkglist_whatever. Look inside /live and you should see initrd.img vmlinuz and filesystem.squashfs. Inside /isolinux should be live.cfg (the boot menu) and a bunch of other files that are the same ones in /usr/lib/refractasnapshot/iso/isolinux/.
Edit: you could do the same with the mounted usb and look inside the directories.
ls /media/computerbob/liveiso/live # and so on.
The button is called "Edit".
rolfie
And if you are logged in, there's a link that says 'Mark SOLVED' just below the edit button and next to the link that says 'Post reply'.