Before I updated the kernel, it all worked fine connecting to the internet etc.
After the update, no access to the internet. I rolled back the kernel to the previous, still no luck.
I am using win 10 to type this. I have tried a few things including checking the cables, changed the cables, made sure the settings in yast haven’t changed, check the permissions and yet it still does not connect. Under yast I clicked on hardware for a list and checked that the module was loaded and it is. I am loathe to re-install or use the upgrade/downgrade option as I have finally got my scanner to work by blind luck.
That is some vague information. What kernel update from what to what? Is that kernel update just part of the regular updates of Leap 15.0 or is it some special kernel?
And then you should of course gather information from your system. To begin with to see what NICs you have:
It is vague as I cannot boot into KDE or any DE, the screen is flickering badly and when I press a key it doesn’t always register as I am trying to enter root and the password so that I can re-install the Nvidia drivers and then I can give you the information
Of the 20 machines I upgraded to 15.0 - I had 2 that would not connect - the /etc/resolv.conf file was corrupt. This is what I put in to fix it. I added the google DNS servers.
> cat /etc/resolv.conf
/etc/resolv.conf file autogenerated by netconfig!
Before you change this file manually, consider to define the
static DNS configuration using the following variables in the
/etc/sysconfig/network/config file:
NETCONFIG_DNS_STATIC_SEARCHLIST
NETCONFIG_DNS_STATIC_SERVERS
NETCONFIG_DNS_FORWARDER
or disable DNS configuration updates via netconfig by setting:
NETCONFIG_DNS_POLICY=’’
See also the netconfig(8) manual page and other documentation.
Note: Manual change of this file disables netconfig too, but
may get lost when this file contains comments or empty lines
only, the netconfig settings are same with settings in this
file and in case of a “netconfig update -f” call.
Please remove (at least) this line when you modify the file!
I can imagine someone doing literally what that line says and remove the comment.
Of course, that would do no good,
The line should actually instruct to remove the lines <following> which are the IP addresses of the nameservers.
Once those lines are removed (or the entire /etc/resolv.conf file deleted), the nameserver configuration wll be re-generated by either rebooting or restarting the network service.
All working now, for some reason when I upgraded to kernel 4.12-lp150.12.13 etc it would connect. How I solved it, apart from some swearing>:).
Boot up from dvd, select upgrade and select kernel, the previous one. Reboot. I then had to enter as root in the console and re-install the nvidia drivers which was not easy as the screen was flickering badly and it was difficult to enter the details as the system did not always recognise the keys being pressed.
You might take a look into whatever method you’re using to install your nVidia drivers, whether there is a dkms option.
dkms is an option where when it’s part of the module installation, the module will automatically update whenever the kernel is upgraded.
Otherwise, as you’ve discovered…
Manually built and configured kernel modules are built to a specific kernel, and when the kernel changes will break the module you installed.
IMHO your interpretation is wrong. This comment line is significant for the network initialization scripts (wicked and NM). Being removed it signals that the system manager configured manually and that thus the file should not be changed with data obtained from DHCP.
The “at least” suggests that when the system manager configured himself, the comment is not really applicable anymore and one might want to remove it completely or replace it with ones own comment.
I don’t use Wicked or Yast - real Unix people don’t need crutches, we use the command line - I wanted the that file contents to remain the same in case I want to get off Google DNS and Network Manager does fine with it in. I also manage my firewall with iptables and ipset to ban the 1000’s of login attempts to the one Linux box that I can access from outside my home. Just look in /var/log/audit/audit.log to see who is attacking your machine. I even move the ssh port but that does not deter the determined. I also ban IP’s from countries that are not need to maintain OpenSUSE.
I am an old Unix person - I wrote a TCP/IP stack back in the 80’s not a full stack but enough to let CP/M telnet, yp, and ftp and it fit in a 64K CP/M S100 machine with a thick ethernet adapter - all in Z80 assembler - I even wrote drivers for Unix when they switched to streams (System V) - trying hard to see why so many changes are needed that worked fine and reliably for decades. Linux used to work fine in 128K in 32 bit mode - now it needs gigs of ram and 64 bits - sad that 99% of the users don’t need all of that. Heck the browsers now need a gig to work. Early Netscape fit in 64K but crashed every other page.
But here you are advising people that you can pretty sure that they use either wicked through YaST (or through editing the file in /etc/network) or NetworkManager.
So either tell them that this advise is for “real Unix people” that “don’t need crutches” or better refrain from giving help.