Discuss January 2021

discuss@lists.centos.org

95 participants
46 discussions

KVM support on RT Kernerl
by Danishka Navin 21 Jan '21

21 Jan '21

Hi, I have deployed CentOS 7.6 and then installed RT kernel. Later KVM installed on the same box. I noticed that KVM module was not loaded and then switched to the generic kernel. KVM module was loaded in the generic kernel. Is this a known issue? I would like to use KVM with RT kernel. Regards, -- Danishka Navin

1 0

Compiling Qt applications in docker
by H 21 Jan '21

21 Jan '21

I am working on compiling an application using Qt and am a newbie. I already have Qt on my CentOS 7 system but this application requires a later version of Qt not available for C 7. It also requires postgreSQL libraries and I am using the latest version of that, ie pgsql 13. I have successfully compiled it in a docker environment and I can launch it on the desktop. I have, however, not tested it against postgreSQL on my computer - yet. I do not want to run it in docker though but on my "regular" computer outside docker, as well as be able to move it to other computers I have. I am only using docker as a sandbox while working on compiling etc. I am not sure how I should set it up - outside docker - so that the Qt libraries it needs do not foul up the rest of my system but are only seen by this particular application. Can anyone offer a suggestion? Should the compiled application eventually reside in its own directory under /usr/local/bin and where should the Qt libraries be? In the same directory as the app? If not, where should they be placed? I tried to compile it to use the Qt libraries statically linked but the app does not run that way in docker but crashes with a segmentation fault. Suggestions/best practices appreciated!

1 0

mcelog service error flagged in Cockpit
by Robert G. (Doc) Savage 20 Jan '21

20 Jan '21

Whenever I log into Cockpit, it flags mcelog as a service failure-to- start failure. systemctl confirms this. # systemctl status mcelog * mcelog.service - Machine Check Exception Logging Daemon Loaded: loaded (/usr/lib/systemd/system/mcelog.service; disabled; vendor preset: enabled) Active: failed (Result: exit-code) since Sun 2021-01-10 18:06:36 CST; 1 weeks 2 days ago Main PID: 21959 (code=exited, status=1/FAILURE) Jan 10 18:06:36 lionstore.protogeek.org systemd[1]: Started Machine Check Exception Logging Daemon. Jan 10 18:06:36 lionstore.protogeek.org mcelog[21959]: mcelog: ERROR: AMD Processor family 23: mcelog does not support this processor. Please use the edac_mce_amd module instead. Jan 10 18:06:36 lionstore.protogeek.org mcelog[21959]: CPU is unsupported Jan 10 18:06:36 lionstore.protogeek.org systemd[1]: mcelog.service: Main process exited, code=exited, status=1/> Jan 10 18:06:36 lionstore.protogeek.org systemd[1]: mcelog.service: Failed with result 'exit-code'. A search through BZ indicates mcelog hasn't been supported on AMD processors s since family 21. How do I tell Cockpit to use this alternative module? Perhaps more importantly, why doesn't Cockpit automatically set up this module when it's installed on an AMD system? --Doc Savage Fairview Heights, IL

1 0

btrfs to ext4
by Christopher Wensink 19 Jan '21

19 Jan '21

Hello CentOS Gurus, I have just learned that our backup system natively supports restoring vmware VM individual files from a vmware based backup for ext3,ext4 file systems but not BTRFS and some of our VM's run BTRFS as the root file system. Does anyone know if it's possible to convert BTRFS partitions to ext4? Chris -- Christopher Wensink IS Administrator Five Star Plastics, Inc 1339 Continental Drive Eau Claire, WI 54701 Office: 715-831-1682 Mobile: 715-563-3112 Fax: 715-831-6075 cwensink(a)five-star-plastics.com www.five-star-plastics.com

2 1

Restore pristine SELinux configuration ?
by Nicolas Kovacs 19 Jan '21

19 Jan '21

Hi, I have CentOS 7 running on a public server hosting all sorts of web applications, mail, XMPP, MPD, etc. How do I reset SELinux configuration to defaults? I know how to reset all my custom booleans to the initial state. # cat /etc/selinux/targeted/active/booleans.local # This file is auto-generated by libsemanage # Do not edit directly. httpd_unified=1 httpd_can_sendmail=1 spamd_enable_home_dirs=1 httpd_can_network_connect=1 ftpd_full_access=1 mpd_enable_homedirs=1 named_write_master_zones=1 Starting from there, I can manually reset them to 0 with setsebool. On the other hand, I don't know how I would do something similar with the SELinux modules. I vaguely remember having created some of these, for example for Fail2ban to work correctly. But I don't remember what I did here over the years, what modules I created, etc. How would I recreate the default SELinux configuration without having to wipe and reinstall the whole server? Cheers, Niki -- Microlinux - Solutions informatiques durables 7, place de l'église - 30730 Montpezat Site : https://www.microlinux.fr Blog : https://blog.microlinux.fr Mail : info(a)microlinux.fr Tél. : 04 66 63 10 32 Mob. : 06 51 80 12 12

2 2

Kickstart "Error setting up base repository"
by Alex Kirk 19 Jan '21

19 Jan '21

I recently built my first CentOS kickstart config on a CentOS 7 system; it runs well, but the OS and its tools are old enough that it's causing pain with the apps I want to run on top of said system, so I'm trying to port it to CentOS 8. I've taken the anaconda-ks.cfg file created from a successful manual install (of both regular 8 and Stream), placed it on my DVD ISO, and wrapped things back up like I was doing with CentOS 7. I'm successfully beginning the installation, but running into "Error setting up base repository" - despite having not touched either the filesystem structure of the ISO, or the generated kickstart file that points at my repos. Here is the relevant config: #version=RHEL8 # Use graphical install graphical repo --name="AppStream" --baseurl=file:///run/install/sources/mount-0000-cdrom/AppStream %packages @^minimal-environment @development kexec-tools %end # Keyboard layouts keyboard --xlayouts='us' # System language lang en_US.UTF-8 # Network information network --hostname=localhost.localdomain # Use CDROM installation media cdrom # Run the Setup Agent on first boot firstboot --enable ignoredisk --only-use=sda autopart # Partition clearing information clearpart --none --initlabel # System timezone timezone America/New_York --isUtc %addon com_redhat_kdump --enable --reserve-mb='auto' %end %anaconda pwpolicy root --minlen=6 --minquality=1 --notstrict --nochanges --notempty pwpolicy user --minlen=6 --minquality=1 --notstrict --nochanges --emptyok pwpolicy luks --minlen=6 --minquality=1 --notstrict --nochanges --notempty %end When I switched over to a console to try to troubleshoot this, I noticed that /run/install/sources/mount-0000-cdrom/AppStream didn't exist (the "mount-0000-cdrom" directory is empty), but /run/install/repo/AppStream does, and appears to have all the right information within it. I've tried specifying that directory as the baseurl instead, and get no difference in behavior. The only other file I've touched is isolinux/isolinux.cfg on the DVD, with the relevant line now being: append initrd=initrd.img inst.ks=cdrom:/ks.cfg inst.stage2=hd:LABEL=CentOS-Stream-8-x86_64-dvd quiet What am I doing wrong here? This seems like basic config, but obviously I've missed something... -- Alex Kirk Global Principal, Suricata 404-291-6588 akirk(a)corelight.com

3 4

centos-release-gluster8 for centos 7
by Nicolas Zuber 19 Jan '21

19 Jan '21

Hi all, We have some Centos 7 server running with gluster 7 installed. We are using the the packages provided by the storage SIG (centos-release-gluster7). "Yum search centos-release-gluster" shows centos-release-gluster7 as the latest gluster version provided. But according to https://wiki.centos.org/SpecialInterestGroup/Storage there should be already packages for the gluster 8 release available. So I am wondering what I can do to be able to install glusterfs 8 from the storage SIG. Thanks, Nicolas

2 2

How to do virtual IP on NetworkManager
by Jerry Geis 18 Jan '21

18 Jan '21

Hi All, I use virtual IP from time to time on CentOS 7. just take something like eth0 and make eth0:1 put in there the IP and subnet and bring up the new IP and it works. How do I do that with NetworkManager on the command line (assuming all static IP information) I do something like this from nmcli to set the static address: nmcli connection modify eth0 ipv4.method manual ipv6.method ignore autoconnect yes ipv4.addr 192.168.1.8/24 gw4 192.168.1.1 ipv4.dns 8.8.8.8 How then do I add a virtual IP? Thanks Jerry

4 4

Graphic setup issues...help wanted
by Kay Schenk 16 Jan '21

16 Jan '21

The problem. I can NOT change my screen resolution from GUI. I boot to multi-user, then get "into" graphical. First, some background. My original install of CentOS 8 -- maybe 8.2? -- about 3 months ago -- went pretty well but I kept having lockup issues. No issues with graphics using intel graphics or sound using either aplay at command line or GDM. (I also have NVIDIA which I have not configured.) I upgraded to 8.3 at some point. Still all OK with graphics but still lock up issues. At some point, I decided to start completely over and did a clean install of 8.4. I am now using CentOS8 Stream as my primary repo. Some questions? When I first boot, where is the probe of my video card done? Looking at DMESG, I couldn't determine this. When I "start" graphical, I get the following messages which are apparently related, from Xorg.1.log module loading section... [ 985.857] (II) LoadModule: "glx" [ 985.857] (II) Loading /usr/lib64/xorg/modules/extensions/libglx.so [ 985.859] (II) Module glx: vendor="X.Org Foundation" [ 985.859] compiled for 1.20.10, module version = 1.0.0 [ 985.859] ABI class: X.Org Server Extension, version 10.0 [ 985.859] (==) Matched modesetting as autoconfigured driver 0 [ 985.859] (==) Matched fbdev as autoconfigured driver 1 [ 985.859] (==) Matched vesa as autoconfigured driver 2 [ 985.859] (==) Assigned the driver to the xf86ConfigLayout [ 985.859] (II) LoadModule: "modesetting" [ 985.859] (II) Loading /usr/lib64/xorg/modules/drivers/modesetting_drv.so [ 985.859] (II) Module modesetting: vendor="X.Org Foundation" [ 985.859] compiled for 1.20.10, module version = 1.20.10 [ 985.859] Module class: X.Org Video Driver [ 985.859] ABI class: X.Org Video Driver, version 24.1 [ 985.859] (II) LoadModule: "fbdev" [ 985.859] (II) Loading /usr/lib64/xorg/modules/drivers/fbdev_drv.so [ 985.859] (II) Module fbdev: vendor="X.Org Foundation" [ 985.859] compiled for 1.20.3, module version = 0.5.0 [ 985.859] Module class: X.Org Video Driver [ 985.859] ABI class: X.Org Video Driver, version 24.0 [ 985.859] (II) LoadModule: "vesa" [ 985.859] (II) Loading /usr/lib64/xorg/modules/drivers/vesa_drv.so [ 985.859] (II) Module vesa: vendor="X.Org Foundation" [ 985.859] compiled for 1.20.3, module version = 2.4.0 [ 985.859] Module class: X.Org Video Driver [ 985.859] ABI class: X.Org Video Driver, version 24.0 [ 985.859] (II) modesetting: Driver for Modesetting Kernel Drivers: kms [ 985.859] (II) FBDEV: driver for framebuffer: fbdev [ 985.859] (II) VESA: driver for VESA chipsets: vesa [ 985.859] (EE) open /dev/dri/card0: No such file or directory [ 985.859] (WW) Falling back to old probe method for modesetting [ 985.859] (EE) open /dev/dri/card0: No such file or directory From messages I've seen on this problem, the lack of /dev/dri/card0 is the crux of the issue but I have NO idea of how it gets created. So, right now I'm stuck at a default resolution of 1024 x 768 Thanks for any help. -- ------------------------------------- "Don't let anyone dull your sparkle." MzK

2 3

bnxt_en :: interface renamed with a suffix
by Adrian Sevcenco 16 Jan '21

16 Jan '21

Hi! i had a very unpleasant surprise when an unconnected interface named enp97s0f1 was renamed when connected .. and after reboot i seen this in dmesg: [ 3.665192] bnxt_en 0000:61:00.1 enp97s0f1d1: renamed from enp97s0f1 does anyone have any idea what is going on? I disabled anything i could (to have a proper normal nic that do not act and do things behind the OS) like making sure that the broadcom nic is set to SF and disabled any function that could be disabled.. so why the rename? Thanks! Adrian

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Discuss January 2021