You are not logged in.
This might be a useful post:
That's the default setting now.
I'd like to see /var/log/refractasnapshot.log. Send it to me in email or paste it somewhere and give me the link. Or read it and see if there's something obvious. I wouldn't expect the decompression error on the initrd to cause a 0-byte iso to be created.
Edit: I just realized you ran the non-gui version, which doesn't make much of a log file unless you run it with the debug option. (That happens automatically with the gui version.) Run sudo refractasnapshot -d
Make sure you didn't run out of hard drive space. You might need to exclude some things to make the system copy and the squashfs smaller. (3.9G is pretty big.)
I've been running beowulf with static IP set in /etc/network/interfaces since the beginning of the year. Configure it in the interfaces file or use a network manager (like wicd) but don't do both - they may fight with each other.
sgage: volumeicon-alsa if you can't find what you're looking for. (I assume you aren't using pulseaudio, which has its own volume control - pavucontrol)
ralph.ronnquist wrote:Beowulf beta "netinstall" without network gives you nothing. Try using "server" or "desktop" rather.
Not sure i understand what you mean here? I have network connection, that being usb0?
In the past, netinstall isos would give you a small and functional base system if you installed without a network. The current netinstall isos won't do that. For an offline install, people should use the server or desktop installer isos or the live isos.
dev-1-dash-1: Thanks for the feedback. We have our wizards working on it.
"device /dev/sd<x> not available in udev database even after waiting for 1000000 microseconds.<another timeout>"
I've gotten this message trying to install grub in a chroot. It goes away in a couple minutes. I'm unable to reproduce it by booting the live-iso from usb.
dev-1-dash-1: I don't know if you can use jessie to debootstrap beowulf. I would avoid using 'testing' because debian and devuan testing are not the same right now. It might cause problems.
I also don't know if you can install beowulf's debootstrap in jessie. Can't hurt to try.
I do know that all the devuan live isos have debootstrap installed. You could do it from a live session.
kapqp: I see only two lsb packages - lsb-release and lsb-base. They both get installed with the default desktop. The minimal-live just has lsb-base.
Yes, you can create your own live-iso with refractasnapshot, and then install from the live media with refractainstaller.
Documentation is here:
https://refracta.org/documents.html
Make sure build-essential is installed. (Yes, the package name is singular.)
gvfs, gvfs-common, gvfs-daemons and gvfs-libs should all have been installed with the desktop.
gvfs-backends does not get installed automatically. That's just the way the dependencies are set up. I could easily add that to the desktop-live, but adding it to the default netinstall would be a lot more work.
Stable release will happen when we're satisfied that all the installer media are good. If you're wondering when to upgrade from ascii, you don't need to wait. All of our forked packages for beowulf are done.
lsb-release does get included in the default desktop install.
And for those who care:
apt purge pulseaudio avahi-daemon
apt autoremoveis easy.
I think I saw the missing hostname issue once. Didn't happen with most of the installs I did.
Size of the iso was a concern, and the graphical installer takes up a lot of space.
I think you're right about nano missing. I noticed that in testing some earlier builds, and I forgot about it because I was troubleshooting something else. We should be able to add that.
The default setup results in the grub menu showing 'Debian' instead of 'Devuan'.
If you're using legacy bios boot, you can just change /etc/os-release to show ID=devuan and then run update-grub
If you're booting uefi, you'll also need to remove grub-efi-amd64-signed and just use grub-efi-amd64. You will not be able to use secure boot this way.
Stop calling it UTC time. That's not what it is. It's local time.
The very first code block in your first post is correct. The hwclock command is doing exactly what the man page says it does. It's showing you the local time and your local offset from UTC in both cases.
One shows the correct local time because you gave the hwclock accurate information, and the other shows the incorrect local time because you gave the hwclock command incorrect information.
hwclock --utc
2020-02-29 02:15:50.889268+03:00
hwclock --localtime
2020-02-28 23:16:54.311448+03:00You're welcome to contact the author(s) of hwclock and ask them to change the behavior of the command, but I doubt they'll change it.
If you didn't get the 4.19 kernel, you probably don't have the kernel metapackage:
apt install linux-image-amd64Then you will always get the latest kernel for the system.
If you want to see what kernels are available:
apt-cache search linux-imageWe built beowulf installer images for armel, armhf and ppc64el. If you have appropriate hardware, please test and report.
armel (no mini.iso for these. I hope you know what to do.)
https://pkgmaster.devuan.org/devuan/dis … nt/images/
kirkwood/netboot/marvell/{dreamplug,guruplug,sheevaplug,sheevaplug-esata}
kirkwood/netboot/seagate/dockstar
orion/netboot/buffalo
armhf
https://pkgmaster.devuan.org/devuan/dis … t/mini.iso
ppc64el (not arm, but it's on my list.)
https://pkgmaster.devuan.org/devuan/dis … t/mini.iso
Thanks,
fsmithred
I didn't think you would need to use mknod. Since there was no actual device for /dev/sdc, I would expect that there was no /dev/sdc in the filesystem until you made it. I only see as many devices listed as actually exist on my systems. Those get created by udev at boot or when you plug in a removable drive.
I think you should be able to delete both of those without any problems. Maybe reboot, but that's probably not necessary.
I guess you now have a file at /dev/sdc that holds the contents of "a-linux-distro". If there is no disk /dev/sdc, you can delete the file.
Do you know when and where this change took place? Was it in debian or upstream?
Where is '--tcp-option' from? It is not mentioned in man iptables.
rolfie, I got results similar to yours, except I'm the other side of UTC, so mine have '-' instead of '+'. Don't feel bad about being confused. I've been confused by hwclock for years, so I stay away from it. I set it to UTC and use ntp or ntpdate to keep the right time.
The driver must be in the netinstall iso, or I would not have been able to get online. It's possible that the driver in ascii isn't new enough for your new hardware. In that case, you could try some live isos to see which kernel works. If the ascii 2.1 desktop-live doesn't work, you can try it with a backports kernel. I made one for that purpose:
https://get.refracta.org/files/experime … 5_1512.iso
Note: devuan debootstrap is included in all the live isos.
The e1000e driver is in my ascii - it's in the kernel configs for 4.9.0-9 and 4.9.0-11. I don't know why you folks aren't seeing it. On my Thinkpad T420, lsmod shows that it's loaded.
If I boot devuan_ascii_2.1_amd64_netinstall.iso I get an ethernet connection on the Thinkpad, and lsmod shows that e1000e is loaded. This installer iso has 4.9.0-11.
While you're at it, you might want to install the 4.19.0-8 kernel from beowulf. Or install linux-image-amd64 and you'll always get the newest kernel for your system.
Make sure os-prober is installed and run update-grub and it should create boot entries for the other linux systems.
Note: If the other systems are encrypted, update-grub will not see them. In that case, create menuentries for them in /etc/grub.d/40_custom, and then run update-grub.
Reportbug is partially broken. I believe you can use it to generate a report, but you have to email the report manually.
To submit a new bug report, send email to submit@bugs.devuan.org with a descriptive subject line and the first line of the body should be
Package: <package name>
Follow-up messages go to <number>@bugs.devuan.org