You are not logged in.
Hello:
Head_on_a_Stick wrote:@OP: create a file at /etc/sysctl.d/reisub.conf ...
Yes.
Creating the file in my box made it work.And yes, it did work before without the /etc/sysctl.d/reisub.conf file.
Now, how long ago was before, I have not had a chance to use it much, so I really can't say.
O.
echo "kernel.sysrq=1" >> /etc/sysctl.confThat's in my script to make refracta-8, which was ascii, but not for 7, which was jessie. So it started in debian stretch.
And I quit using -e and -i because all they do is generate a console message saying that they are disabled.
For the initial transfer, maybe you can bypass the router and connect one box to the other. You'd have to set static IP adresses and use the address on the remote box as the gateway on the local box. Since they're both gigabit connections, you don't need a crossover cable. A regular one will do.
Sometimes rebooting a router will help if it's been doing a lot of work.
If you do a lot of local transfers, you could add a faster router to the slower one and keep your local net on the faster one.
Not sure about how the phone works, but it might be possible to set the modem-router to bridged mode so it acts like a plain modem and use your own router off of that.
I'm getting 11MB/s with rsync or sftp on the local network, Cat6 with gigabit router.
USB is faster, but that requires getting out of the chair.
Do this first!
apt install wicdIt's a transitional metapackage that will pull in network-manager-gnome during the upgrade.
If you would rather have connman instead of network-manager-gnome, install connman before the upgrade.
Without that you will be left without network. If you get stuck without network, run
dhclient eth0to get a connection.
Refracta uses only Devuan repositories. Refracta IS Devuan, just with a different name and some different configurations. Your boot menu might say Devuan after the upgrade. You can change that in /etc/default/grub if you want.
Use upgrade instructions for Devuan. Do not use instructions for migration from Debian.
Refracta tracks Debian version numbers because it started out as a Debian re-spin with Debian Lenny (5). It was switched to Devuan with Jessie, and it didn't make sense to go from version 7 to version 1.
Ah, I did just think of one package that isn't in the repo. refracta2usb. The newest version of that is here: https://sourceforge.net/projects/refrac … -2.4.3.deb
It will work with beowulf, chimaera, and probably daedalus, too.
I'm not sure I understand what you want to do. Maybe you think it's more complicated than it is.
Change "beowulf" to "chimaera" in sources.list, then 'apt upgrade' and 'apt dist-upgrade' the same as you would with pure devuan. You do not need to reinstall.
sed -i 's/beowulf/chimaera'/g /etc/apt/sources.list
apt update
apt upgrade
apt dist-upgradeYou will be asked about some config files that have been changed. Say 'yes' to take the new versions of the config files.
There's no support for lvm in refractainstaller. If you do full disk encryption, you would need to enter the luks passphrase once for grub and then once for the root partition. In that case, /boot is in the root partition. If you made a separate partition for /home, you would need to enter the passphrase for that partition, too.
The only filesystem that refractainstaller knows about is ext[2,3,4].
OK, now here's the part where I tell you about going past the limits.
You could make luks keyfiles to avoid or reduce passphrase entries after the install.
You can use other filesystems if you use the cli installer and set the config file to no_format=yes. Partition and format as you want, then run the installer. You might need to install something in the live system to do this. I don't know what xfs needs.
You can also use the cli installer and manually create lvm or raid. I left some notes about that here: https://dev1galaxy.org/viewtopic.php?id=2323 One thing that's changed is that you don't have to end the /dev/mapper/name with a digit.
ForgeOS (Formerly Michael's Devuan Edition) uses a modified version of the cli refractainstaller that does support lvm. I've only had a couple of brief looks at it and haven't had a chance to dive into the code. The other notable difference is that you make all your selections at the beginning instead of waiting for questions to pop up.
Refractainstaller has supported full disk encryption for at least a couple of years. Select encryption for the root partition and do not select a separate /boot partition.
Glad you got it fixed. For future reference, those GDBus errors complaining about missing .service files are just noise from programs that expect you to be running systemd, even though they may not require it. I see them often when I start programs from a terminal.
Have you tried logging out of the desktop and then log in again? I've deleted my panel a few times by accident. It always comes back on login.
If you're in the middle of something and can't log out, run xfce4-panel from the Run dialogue or from a terminal. You can get a Run dialogue with alt-F2.
The watermark is barely visible in the gray field. It's the round peppermint candy. (And now I want one. Or a candy cane, but they're all gone.)
I got a brief answer from Jaromil: "yes why not"
What does the logo for the debian build look like? I tried to find it, but failed. Or maybe the examples I saw were too small.
I bumped the request up to our traveling guru at dyne.org. Be patient. He moves faster than email.
bluetoothd version 5.62 maybe needs to be debugged and repaired.
Yeah, maybe. Some of these bugs go back to 2009. I can't tell from this page if they're unfixed or if they just don't ever close their bug reports.
https://bugs.debian.org/cgi-bin/pkgrepo … kage=bluez
Thanks for the info. Someone else may run into the same problem.
The *.d directories are a place for the administrator to put customized configurations that will not be wiped out if a new version of the config file is installed/upgraded. Something is programmed to check for files in that directory, and it's letting you know that there's nothing there.
It should be safe to ignore. I don't have that directory on my chimaera or daedalus installations.
My syslog looks like #2. It seems that a lot of my entries for going into sleep have the same timestamp as the entries coming out of sleep. Not sure what to make of that. Or of your #3 paste.
During my testing, I did have occasion to not wake up from sleep. I had other problems after reboot, like not being able to drop to console with ctrl-alt-F2. I ran fsck from another system on the same machine and it fixed some orphaned inodes. That's probably not your problem, but might be worth trying fsck from a live-usb.
Have you tried running loginctl suspend to see if it acts differently? On mine it does the same as pm-suspend.
There's no metapackage for backports kernels like there is for the main kernel. The only updates you would get would be if the kernel version didn't change. I don't know if that happens with backports kernels.
Thanks, nili. I should remember to search the page for words. 7G is more than enough, assuming it's all working.
The only time I ever had a cpu running at 11C was with the computer on a hardwood floor, in the winter with sub-freezing temperatures, and the windows in the apartment below me were open. My floor was very cold.
I vote for staying in this thread. These things may all be related, and someone else might run into similar problems.
Is it a laptop? (Do we need to deal with lid switch and power-manager?)
Look in /var/log/pm-suspend.log to see information related to entering and leaving suspend state.
Look in /var/log/syslog for info about leaving suspend state. Find where it says PM: suspend entry (deep) to see what's happening when you leave suspend. I could not find anytihing in syslog to show when it went into suspend. (And I started with an empty syslog, so there's no way I missed it.)
How much memory does it have? If it's in the inxi output, I'm missing it.
There's a big difference in mem requirements between jessie and chimaera. It keeps going up with each release. Run memtest86+ overnight to see if there's bad RAM.
If you're still having trouble with suspend after you get the packages all straightened out, you might take a look at this page: https://wiki.ubuntu.com/DebuggingKernelSuspend
Someone in IRC gave me the link. I haven't had a chance to read it, and I will be busy with other things today.
consolekit and libpolkit-gobject-consolekit-1-0 should get removed when you install elogind and libpam-elogind. And libpolkit-gobject-elogind-1-0 will be installed automatically.
If you use apt or apt-get to install elogind and the others, it will put a lot of packages on the autoremove list, but it won't actually remove them. I'm pretty sure once you do that and libelogind0 is in the system, those packages will be taken off the autoremove list.
To see the autoremove list without removing anything.
apt -s autoremoveIf it doesn't clear after libelogind0 is installed, you can reinstall consolekit, libpam-ck-connector and libelogind0. Then we can talk about setting up sudo for only shutdown/reboot.
If you use aptitude to install, then the packages will be automatically removed during the install. Don't do that here.
If that's the output of apt-get install libelogind0 then I have no idea what's going on. That should not happen. Try installing all at once and see if that works better.
apt install elogind libpam-elogind libelogind0or maybe
aptitude install elogind libpam-elogind libelogind0and see if aptitude gives you some other solutions.
I'm glad you waited. Install libelogind0 first to replace libsystemd0, and that should prevent all those removals from happening. Then install elogind and the others.