The small linux problem thread

Yes the patching will save the data. It might look unresponsive for a while, the firmware update software isn't very verbose where it counts and might take its time, but it also rearranges data, it doesn't just patch the firmware. So just let it do its thing and it'll be fine. I think that will solve the problem, it did when I had it.

Just gave it a go, drive shows up as "unknown drive" in the Samsung "Magician" software. Is there something that will try flashing even if it can't ID the drive? I've only got two drives in the windows box, so I'm confident I know which is which.

I've not had that problem I think (it's a few years ago lol). What I do remember is that there was quite an extensive documentation on those errors on Samsungs website. Samsung are quite specific about what causes what because their user base is so big, they have info on most issues.

One thing I can think of, is that Samsung drives have the functionality of linking with BIOS encryption functionality. If that would be active, the drive would not be recognisable on a different machine. You would know if that were active because it would mean that a HDD password were set on the PC the disk comes from.

Another thing to remember is that Windows does not recognise linux partitions using linux filesystems.

And another thing is that the PC you've put the ssd into, does not have a UEFI whereas the original PC does, so maybe GPT is not supported and thus the disk is not recognised.

Hoping not to cause a language war - focus on ruby or python for Linux sysadmining? Currently using bash and perl largely.

Also, is brushing up on advanced awk a waste of time?

K thx bai

Question: what is Gruvbox? is it just a plugin for vim, or is it something i can apply to my entire terminal emulator?

Depends on the kind of sysadmin duties. A lot of tools in the field of provisioning, load management, distributed assets, etc are based on python. Python is a very handy tool because it is easy to use and there are interfaces for all major subsystems. I would suggest you would look at the packages on your admin system, look at what subsystems you use most, and whether there are python tools for it that exist already, that you can use as they are or easily adapt to your specifications, or whether there is nothing and maybe you like something else more than Python.

1 Like

I have 2 CentOS machines.

1 CentOS VM running CentOS 7 and Spacewalk 2.6

1 CentOS physical box on the same network running CentOS 6.8

I was following this walkthrough https://github.com/spacewalkproject/spacewalk/wiki/RegisteringClients

I get to the point where you input the SSL certificate on the client
# rpm -Uvh http://YourSpacewalk.example.com/pub/rhn-org-trusted-ssl-cert-1.0-1.noarch.rpm
it imports and gives a warning: The dynamic CA configuration feature is in the disabled state.

Then I run command found in number 4

# rhnreg_ks --serverUrl=https://YourSpacewalk.example.org/XMLRPC --sslCACert=/usr/share/rhn/RHN-ORG-TRUSTED-SSL-CERT --activationkey=<key-for-channel>

It errors saying

The SSL certificate failed verification.

Any tips? couldn't find much from googling.

EDIT: I found the problem. I forgot that I reinstalled Spacewalk on the server due to the previous install being buggy. The install process creates an SSL certificate. I didn't realize that the new ssl certificate is not created over the old one.

Old: rhn-org-trusted-ssl-cert-1.0-1

New:rhn-org-trusted-ssl-cert-1.0-2

Pulled down rhn-org-trusted-ssl-cert-1.0-2 with the client, ran the rhnreg_ks command and it connected fine!

No wonder google didn't help me, nobody is as stupid as I am :P

Oh this is perfect!

I've got an install of Antergos (XFCE for the DE) in Virtualbox. The clock on the desktop shows the right time, however the clock on the lock screen does not. It's a good 7 hours fast. I've consulted the google and have come up with nothing. I was initially thinking it was a BIOS setting for the VM...but wouldn't that also mess with the normal desktop clock? My thinking is that if NTP was working correct it would also fix the lock screen time too. Maybe I'm wrong?

Funny you should post this, but I was playing with time settings to fix my issue right above you.

iirc: There are 3 different time settings in Linux. The Hardware time, Kernel Time, and System time. The System time gets it from the hardware time very early on in the boot process.

There is also a time adjustment file that is made if the system thinks the time is to fast or slow, Delete this. Then set your time. Where all of your time settings will differ from distros. Then reboot. see if issue is fixed.

You can use

#date MMDDhhmm
#hwclock --systohc

man date

man to use date correctly.
hwclock --systohc will set the system clock according to the hardware clock

or

alternatively

hwclock -ntp --systohc

to also set time to use ntp

Then check /etc/adjtime for the time adjustment file

That was for my centos distro, not sure if itll work for you.

Reference:
https://eloquence.marxmeier.com/sdb/html/954237377.html

1 Like

I'll have to look up the equivalent commands for arch (but I can't imagine they'd be wildly different.)

Thanks man!

I can't access my Arch machine through it's hostname. When I try to connect to it through PuTTY with the hostname I set, it claims "Host does not exist". Yet, if I connect through the IP address, it works perfectly. I would be fine connecting through IP, but I don't particularly want my machine set to a static IP, due to being a laptop, and there's no need for static IP. The output of cat /etc/hostname gives me archlaptop (Imaginative, I know).

Is there anywhere besides /etc/hostname that I need to set anything? Because if not, then something else is wrong. This is a predominantly Windows network, but I don't see why that would affect anything. No domain on the home network (Other than the default Windows WORKGROUP). What would I need to set in this case? Something in /etc/hosts?

EDIT:
My /etc/hosts file:
<ip-address> <hostname.domain.org> <hostname>
127.0.0.1 localhost.localdomain localhost archlaptop
::1 localhost.localdomain localhost archlaptop

sounds to me like the dns doesn't resolve your hostname into an ip adress. but this whole ip thingy is not my strength at all.

stuff to get you in the direction (im not sure though, ip adresses, dns, hosts and all that is still a mysterie to me)

etc/nsswitch.conf 
netstat 
ifconfig
ip

Hey, every time i update my kernel the nvidia driver breaks. Its difficult to provide some feedback on errors right now because i ended up bricking the system trying to figure out whats up.
But this is around the fifth time it happens, i install Fedora 25, add rpm fusion repos, do a dnf update, reboot into the most recent kernel, install the nvidia drivers according to rpm fusions HowTo\nVidia, and it works just fine. Some time passes, a kernel update comes up, i add it, reboot and the driver is dead or malfunctioning. I try to fix it and end up breaking the system.

Is there anything im supposed to do before a kernel update? The card is a GTX760, i add the drivers using dnf install xorg-x11-drv-nvidia akmod-nvidia "kernel-devel-uname-r == $(uname -r)"

Edit: apparently there's this solution, however i couldn't find anything about this service outside of this page. https://rpmfusion.org/Packaging/KernelModules/Akmods but at least i know what the problem is now.

It should in theory update the nvidia driver after a kernel update, you need to make sure you pull the latest kernel-devel as well. If its not always doing it you should be able to reinstall the akmod-nvidia

Another solution is to use the dkms version from negativo17 http://negativo17.org/

1 Like

Another Fedora 25 problem here. After i updated to kernel 4.10 the user alphazero1990 is fu***d. I can't change anything not even the background. Everything stays factory reseted. For example the icons on the left if i remove one ti just pops back up... or if i add one it removes itself. But my other user works fine like 100% operational.

I would love to fix this so that next time it happens i know what to do.

What else updated? This won't have anything to do the with kernel.

I assume your talking about icons in the dock/tray on the left of the screen?

1 Like

that's weird.
sounds like your user doesnt have the permission to make those changes and/or the changes this user makes are being overwritten.

you have no problem logging in right?

cat /etc/passwd | grep alphazero1990

output should look like this:
alphazero1990:x:100x:100x:/home/alphazero1990:/bin/bash

oh, i totally mis-read what your posted :smiley:

if you can't change the icons on the dock or what its called then the problem is connectet to your xdm/gdm/etc.

what Desktop Environment are your using?

I am running gnome. I tride wayland and x11 same thing. yes no problem logging in. I can run programs like VLC, Steam. Stuff like gnome tweak tool won't let me change anything.

The output of the command: alphazero1990:x:1000:1000:alphazero1990:/home/alphazero1990:/bin/bash

okay.

when gnome does such weird stuff its mostly because of some extensions or stuff like that.

what you can do is to enable the debugging of gdm.

vim/nano/pico /etc/gdm/custom.conf

    [debug]
# Uncomment the line below to turn on debugging
Enable=true

if you delete the '#' the debugging is enabled. maybe it gives you any sort of error / warning what is behaving not accordingly.

Done that now. So where are those messages displayed?