Discovery plugin / possible to expect a CentOS8 ethernet naming in a near future?

jfwl · May 6, 2021, 8:37am

Hello,

I use the last 3.7 version. Is it possible to expect a next version that will provide the CentOS8/RHEL8 native ethernet network naming conventions ? Because for now, the names are not the same that after install and I would be very convenient to read the same names.
Regards

lzap · May 7, 2021, 8:32am

Hello,

what ethernet naming conventions are on your mind? If you refer to systemd naming:

https://www.freedesktop.org/software/systemd/man/systemd.net-naming-scheme.html

This has been added into RHEL 7.0 with exception of virtio driver which still use eth0. This was a bug that could not be fixed until RHEL 8.0.

jfwl · May 7, 2021, 3:43pm

Hello,
We use default naming from OS.
Here is an exemple !

Here is once RHEL8 installed :

]# cat /proc/cmdline
BOOT_IMAGE=(hd0,msdos1)/vmlinuz-4.18.0-240.el8.x86_64 root=/dev/mapper/vg00-root ro nofb splash=quiet crashkernel=auto resume=/dev/mapper/vg00-swap rd.lvm.lv=vg00/root rd.lvm.lv=vg00/swap rhgb quiet
[root@obench29s ~]# ip link
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: ens1f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
    link/ether a0:36:9f:f1:ff:20 brd ff:ff:ff:ff:ff:ff
3: eno1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether 34:80:0d:9e:e1:70 brd ff:ff:ff:ff:ff:ff
4: ens1f1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether a0:36:9f:f1:ff:21 brd ff:ff:ff:ff:ff:ff
5: eno2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether 34:80:0d:9e:e1:71 brd ff:ff:ff:ff:ff:ff
6: ens1f2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether a0:36:9f:f1:ff:22 brd ff:ff:ff:ff:ff:ff
7: eno3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether 34:80:0d:9e:e1:72 brd ff:ff:ff:ff:ff:ff
8: ens1f3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether a0:36:9f:f1:ff:23 brd ff:ff:ff:ff:ff:ff
9: eno4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
    link/ether 34:80:0d:9e:e1:73 brd ff:ff:ff:ff:ff:ff
10: idrac: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UNKNOWN mode DEFAULT group default qlen 1000
    link/ether 58:8a:5a:f5:a3:07 brd ff:ff:ff:ff:ff:ff

Regards

jfwl · May 7, 2021, 3:49pm

I’ts like the discovery uses biosdevname=1. But we don’t.

lzap · May 10, 2021, 12:30pm

Discovery does not use biosdevname=1 by default and I still do not understand your concern. Please elaborate.

jfwl · May 10, 2021, 12:43pm

Hello,
It does not but look at the example above. I thought it was quite telling yet.
Discovery presents em1,2,3,4 whereas basic installed CentOS presents eno for instance.

lzap · May 12, 2021, 6:02pm

Oh yeah I missed the block (which I reformatted).

I think you are missing biosdevname package on the installed system. It should be present on the discovery image. Can you check that: em1 interface renamed to eno1 ; p1p1 interface renamed to enp129s0f0 after system reboot - Red Hat Customer Portal

Also do this: dmesg | grep "renamed from" on both systems.

jfwl · May 14, 2021, 8:02am

Hello,
Default RHEL8 grub does not use biosdevname.
And my server has already biosdevname installed.

# rpm -qa biosdev*
biosdevname-0.7.3-2.el8.x86_64

# cat /proc/cmdline
BOOT_IMAGE=(hd0,msdos1)/vmlinuz-4.18.0-240.el8.x86_64 root=/dev/mapper/vg00-root ro nofb splash=quiet crashkernel=auto resume=/dev/mapper/vg00-swap rd.lvm.lv=vg00/root rd.lvm.lv=vg00/swap rhgb quiet

# ip link
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: ens1f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
    link/ether a0:36:9f:f1:ff:20 brd ff:ff:ff:ff:ff:ff
3: ens1f1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether a0:36:9f:f1:ff:21 brd ff:ff:ff:ff:ff:ff
4: eno1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether 34:80:0d:9e:e1:70 brd ff:ff:ff:ff:ff:ff
5: ens1f2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether a0:36:9f:f1:ff:22 brd ff:ff:ff:ff:ff:ff
6: eno2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether 34:80:0d:9e:e1:71 brd ff:ff:ff:ff:ff:ff
7: ens1f3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether a0:36:9f:f1:ff:23 brd ff:ff:ff:ff:ff:ff
8: eno3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
    link/ether 34:80:0d:9e:e1:72 brd ff:ff:ff:ff:ff:ff
9: eno4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
    link/ether 34:80:0d:9e:e1:73 brd ff:ff:ff:ff:ff:ff
10: idrac: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UNKNOWN mode DEFAULT group default qlen 1000
    link/ether 58:8a:5a:f5:a3:07 brd ff:ff:ff:ff:ff:ff

For my understanding, the strange fact is to retrieve biosdevname set in the discovery image…

lzap · May 14, 2021, 8:26am

I edited your post for the second time, please use proper formatting when pasting output.

I do not understand.

I asked RHEL engineers here at Red Hat what would cause such behavior, if you say both discovery and provisioned system have default kernel line options and the package installed, I would really expect to have them both the same NIC names. I do not know what is wrong, I am seeing this for the first time.

What hardware and network card is that?

jfwl · May 14, 2021, 9:20am

DELL R640

“I do not understand”
The naming on my installed system seems normal for me.

I was not expecting to get em interface except with biosdevname enabled.
To check, I’ve set biosdevname into the boot kernel line and here is the result

cat /proc/cmdline

BOOT_IMAGE=(hd0,msdos1)/vmlinuz-4.18.0-240.el8.x86_64 root=/dev/mapper/vg00-root ro nofb splash=quiet crashkernel=auto resume=/dev/mapper/vg00-swap rd.lvm.lv=vg00/root rd.lvm.lv=vg00/swap rhgb quiet biosdevname=1
[root@obench29s ~]# ip link
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: ens1f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether a0:36:9f:f1:ff:20 brd ff:ff:ff:ff:ff:ff
3: p1p2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
link/ether a0:36:9f:f1:ff:21 brd ff:ff:ff:ff:ff:ff
4: em1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
link/ether 34:80:0d:9e:e1:70 brd ff:ff:ff:ff:ff:ff
5: p1p3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
link/ether a0:36:9f:f1:ff:22 brd ff:ff:ff:ff:ff:ff
6: em2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
link/ether 34:80:0d:9e:e1:71 brd ff:ff:ff:ff:ff:ff
7: p1p4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
link/ether a0:36:9f:f1:ff:23 brd ff:ff:ff:ff:ff:ff
8: em3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
link/ether 34:80:0d:9e:e1:72 brd ff:ff:ff:ff:ff:ff
9: em4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 34:80:0d:9e:e1:73 brd ff:ff:ff:ff:ff:ff
10: idrac: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UNKNOWN mode DEFAULT group default qlen 1000
link/ether 58:8a:5a:f5:a3:07 brd ff:ff:ff:ff:ff:ff

So this the mystery why the FDI foreman discovery system (and maybe also the live image of RHEL8 ) use the biosdevname like it would be set.

lzap · May 17, 2021, 5:37am

RHEL engineers replied to me: biosdevname is automatically enabled on Dell hardware at RHEL 7. But it isn’t automatically enabled on Dell hardware at RHEL8.

Unfortunately, I do not have a solution for you other than

github.com/theforeman/foreman-documentation

Note about discovery naming scheme

theforeman:master ← lzap:discovery-naming-scheme

opened 05:33AM - 17 May 21 UTC

lzap

+6 -0

SSIA Cherry-pick into: * [ ] Foreman 2.5 (Satellite 6.10) * [ ] Foreman 2.4 * …[ ] Foreman 2.3 (Satellite 6.9) * [ ] Foreman 2.1 (Satellite 6.8)

More details at

We should be looking into upgrading discovery to EL8 this year, there is a major change there going from livecd-creator to lorax, I would rather switch away from live CD completely to dracut instead.

lzap · May 17, 2021, 2:10pm

One thing which comes to my mind is - why exactly do you need interfaces to be the same? I mean, obviously it’s good for consistency, but technically Foreman deploys MAC addresses to ifcfg configuration files, so even if ifcfg-XXXX is named incorrectly, NetworkManager should pick and configure network correctly.

jfwl · May 17, 2021, 2:20pm

Hello,
Thank you I’ve also checked that but for the operator, and for bonding name guess, It would be better to read the final names.
BTW, is it a big deal to make a new version using version 8 ?
Regards

lzap · May 17, 2021, 2:44pm

I do not know, I’d need to take a look. Technically, it’s a change in the tool that builds the image. But we need then to revisit out release process, jenkins building job, downstream process. What can look like an easy change can actually be pretty challenging.

But even if we have EL8-based discovery, you will have the same problem when you want to install an EL7 system. It would be just the other way around.

Try it out yourself, building FDI is not difficult actually:

Instructions will not work, you need to use livemedia-creator instead. I think we might add support for this tho while continue building EL7 FDI image.

lzap · May 18, 2021, 7:40am

For the record, I tried to build discovery off CentOS8. After runniing bulid-livecd from the git repository which writes new fdi-image.ks I edited it accordingly:

network  --bootproto=dhcp
url --mirrorlist=http://mirrorlist.centos.org/?release=8&arch=$basearch&repo=baseos
repo --name="AppStream" --mirrorlist=http://mirrorlist.centos.org/?release=8&arch=$basearch&repo=appstream
repo --name="epel8" --mirrorlist=https://mirrors.fedoraproject.org/mirrorlist?repo=epel-8&arch=x86_64
repo --name="foreman-el8" --baseurl=http://yum.theforeman.org/releases/nightly/el8/$basearch/
repo --name="foreman-plugins-el8" --baseurl=http://yum.theforeman.org/plugins/nightly/el8/$basearch/
module --name=ruby --stream=2.7
#auth --useshadow --enablemd5
clearpart --all --initlabel
part / --fstype="ext4" --size=4000
shutdown

And in the packages section, I added dracut-live but removed OpenIPMI-tools, acpid, bind-utils, isomd5sum, tcpdump, tftp, uuid. Finally I renamed tfm-rubygem-facter to just rubygem-facter. Then the command was:

sudo livemedia-creator --ks fdi-image.ks --make-iso --iso centos.iso --tmp /tmp
2021-05-18 09:19:00,219: livemedia-creator v34.9-1
2021-05-18 09:19:00,219: selinux is enabled and in Permissive mode
2021-05-18 09:19:00,307: disk_img = /tmp/lmc-disk-mr2qmvzx.img
2021-05-18 09:19:00,307: Using disk size of 4002MiB
2021-05-18 09:19:00,307: install_log = /home/lzap/work/foreman-discovery-image/virt-install.log
2021-05-18 09:19:01,053: qemu vnc=127.0.0.1:0
2021-05-18 09:19:01,053: Running qemu
2021-05-18 09:19:01,077: Processing logs from ('127.0.0.1', 43800)
2021-05-18 09:24:22,108: Installation finished without errors.
2021-05-18 09:24:22,109: Shutting down log processing
2021-05-18 09:24:22,109: unmounting the iso
2021-05-18 09:24:22,128: Disk Image install successful
2021-05-18 09:24:22,128: working dir is /tmp/lmc-work-xpjl3jue
2021-05-18 09:24:22,172: Partition mounted on /tmp/tmpt7b80s77 size=4194304000
2021-05-18 09:24:22,172: Using disk size of 4002MiB
2021-05-18 09:24:22,204: [Errno 2] Adresář nebo soubor neexistuje: '/tmp/tmpt7b80s77/boot'

Unfortunately, I am running into an issue I don’t know how to solve. I asked the devs:

github.com/weldr/lorax

Directory not found after successful ISO build

opened 07:34AM - 18 May 21 UTC

closed 03:40PM - 18 May 21 UTC

lzap

Hello, I am upgrading our hardware inventory discovery liveCD (https://github….com/theforeman/foreman-discovery-image) from EL7 to EL8 and lorax, I have performed I hope all the necessary kickstart changes, the ISO image builds fine however livemedia-creator errors out with "Directory or file not found: /tmp". The error message is in Czech locale, I am using /tmp for faster builds (tmpfs) but I did try this on /var/tmp too with the same result: ``` sudo livemedia-creator --ks fdi-image.ks --make-iso --iso centos.iso --tmp /tmp 2021-05-18 09:19:00,219: livemedia-creator v34.9-1 2021-05-18 09:19:00,219: selinux is enabled and in Permissive mode 2021-05-18 09:19:00,307: disk_img = /tmp/lmc-disk-mr2qmvzx.img 2021-05-18 09:19:00,307: Using disk size of 4002MiB 2021-05-18 09:19:00,307: install_log = /home/lzap/work/foreman-discovery-image/virt-install.log 2021-05-18 09:19:01,053: qemu vnc=127.0.0.1:0 2021-05-18 09:19:01,053: Running qemu 2021-05-18 09:19:01,077: Processing logs from ('127.0.0.1', 43800) 2021-05-18 09:24:22,108: Installation finished without errors. 2021-05-18 09:24:22,109: Shutting down log processing 2021-05-18 09:24:22,109: unmounting the iso 2021-05-18 09:24:22,128: Disk Image install successful 2021-05-18 09:24:22,128: working dir is /tmp/lmc-work-xpjl3jue 2021-05-18 09:24:22,172: Partition mounted on /tmp/tmpt7b80s77 size=4194304000 2021-05-18 09:24:22,172: Using disk size of 4002MiB 2021-05-18 09:24:22,204: [Errno 2] Adresář nebo soubor neexistuje: '/tmp/tmpt7b80s77/boot' ``` The `centos.iso` symlink is a CentOS 8.3 full installation DVD. Kickstart is here (we do bit of injecting of some files and minimizing the image to the smallest possible size for PXE): https://gist.github.com/lzap/3924956da758104e9276219ffe5239be Additional logs: ``` 2021-05-18 09:19:00,219 DEBUG livemedia-creator: Namespace(make_iso=True, make_disk=False, make_fsimage=False, make_appliance=False, make_ami=False, make_tar=False, make_tar_disk=False, make_pxe_live=False, make_ostree_live=False, make_oci=False, make_vagrant=False, iso='/home/lzap/work/foreman-discovery-image/centos.iso', iso_only=False, iso_name=None, ks=['/home/lzap/work/foreman-discovery-image/fdi-image.ks'], image_only=False, no_virt=False, proxy=None, anaconda_args=None, armplatform=None, location=None, logfile='/home/lzap/work/foreman-discovery-image/livemedia.log', lorax_templates=None, tmp='/tmp', result_dir=None, domacboot=True, extra_boot_args='', disk_image=None, keep_image=False, fs_image=None, image_name=None, tar_disk_name=None, fs_label='Anaconda', image_size_align=0, image_type=None, qemu_args=[], qcow2=False, compression='xz', compress_args=[], app_name=None, app_template=None, app_file='appliance.xml', ram=2048, vcpus=None, vnc=None, arch=None, kernel_args=None, ovmf_path='/usr/share/edk2/ovmf/', virt_uefi=False, no_kvm=False, with_rng='/dev/random', dracut_conf=None, dracut_args=None, live_rootfs_size=0, live_rootfs_keep_size=False, oci_config=None, oci_runtime=None, vagrant_metadata=None, vagrantfile=None, project='Linux', releasever='34', volid=None, squashfs_only=False, timeout=None) 2021-05-18 09:19:00,219 INFO livemedia-creator: livemedia-creator v34.9-1 2021-05-18 09:19:00,219 INFO pylorax: selinux is enabled and in Permissive mode 2021-05-18 09:19:00,307 INFO pylorax: disk_img = /tmp/lmc-disk-mr2qmvzx.img 2021-05-18 09:19:00,307 INFO pylorax: Using disk size of 4002MiB 2021-05-18 09:19:00,307 INFO pylorax: install_log = /home/lzap/work/foreman-discovery-image/virt-install.log 2021-05-18 09:19:00,308 DEBUG pylorax.imgutils: make tmp mountdir /tmp/lorax.imgutils.nl37yl_1 2021-05-18 09:19:00,984 DEBUG pylorax.imgutils: find /tmp/lmc-ksdir-hxqyz9sc -print0 |cpio --null --quiet -H newc -o | xz -T4 --check=crc32 -9 > /tmp/lmc-ks-q0g822st.img 2021-05-18 09:19:01,053 INFO pylorax: qemu vnc=127.0.0.1:0 2021-05-18 09:19:01,053 INFO pylorax: Running qemu 2021-05-18 09:19:01,053 DEBUG pylorax: ['qemu-system-x86_64', '-no-user-config', '-m', '2048', '-machine', 'accel=kvm', '-kernel', '/tmp/lorax.imgutils.nl37yl_1/isolinux/vmlinuz', '-initrd', '/tmp/lmc-initrd-_wx_h7kw.img', '-drive', 'file=/tmp/lmc-disk-mr2qmvzx.img,cache=unsafe,discard=unmap,format=raw', '-drive', 'file=/home/lzap/work/foreman-discovery-image/centos.iso,media=cdrom,readonly=on', '-append', 'ks=file:/fdi-image.ks inst.stage2=hd:LABEL=CentOS-8-3-2011-x86_64-dvd inst.text inst.cmdline', '-nographic', '-monitor', 'none', '-serial', 'null', '-display', 'vnc=127.0.0.1:0', '-device', 'virtio-serial-pci,id=virtio-serial0', '-device', 'virtserialport,bus=virtio-serial0.0,nr=1,chardev=charchannel0,id=channel0,name=org.fedoraproject.anaconda.log.0', '-chardev', 'socket,id=charchannel0,host=127.0.0.1,port=40401', '-object', 'rng-random,id=virtio-rng0,filename=/dev/random', '-device', 'virtio-rng-pci,rng=virtio-rng0,id=rng0,bus=pci.0,addr=0x9'] 2021-05-18 09:19:01,077 INFO livemedia-creator: Processing logs from ('127.0.0.1', 43800) 2021-05-18 09:24:22,108 INFO pylorax: Installation finished without errors. 2021-05-18 09:24:22,109 INFO livemedia-creator: Shutting down log processing 2021-05-18 09:24:22,109 INFO pylorax: unmounting the iso 2021-05-18 09:24:22,127 DEBUG pylorax.imgutils: remove tmp mountdir /tmp/lorax.imgutils.nl37yl_1 2021-05-18 09:24:22,128 INFO pylorax: Disk Image install successful 2021-05-18 09:24:22,128 INFO pylorax: working dir is /tmp/lmc-work-xpjl3jue 2021-05-18 09:24:22,163 DEBUG pylorax.imgutils: add map loop0p1 (253:8): 0 8192000 linear 7:0 2048 2021-05-18 09:24:22,172 INFO pylorax.imgutils: Partition mounted on /tmp/tmpt7b80s77 size=4194304000 2021-05-18 09:24:22,172 INFO pylorax: Using disk size of 4002MiB 2021-05-18 09:24:22,204 ERROR livemedia-creator: [Errno 2] Adresář nebo soubor neexistuje: '/tmp/tmpt7b80s77/boot' ``` Also: ``` 2021-05-18 09:19:00,308 INFO: Running... mount -o loop /home/lzap/work/foreman-discovery-image/centos.iso /tmp/lorax.imgutils.nl37yl_1 2021-05-18 09:19:00,325 INFO: mount: /tmp/lorax.imgutils.nl37yl_1: WARNING: source write-protected, mounted read-only. 2021-05-18 09:19:00,325 DEBUG: Return code: 0 2021-05-18 09:19:01,053 INFO: Running... qemu-system-x86_64 -no-user-config -m 2048 -machine accel=kvm -kernel /tmp/lorax.imgutils.nl37yl_1/isolinux/vmlinuz -initrd /tmp/lmc-initrd-_wx_h7kw.img -drive file=/tmp/lmc-disk-mr2qmvzx.img,cache=unsafe,discard=unmap,format=raw -drive file=/home/lzap/work/foreman-discovery-image/centos.iso,media=cdrom,readonly=on -append ks=file:/fdi-image.ks inst.stage2=hd:LABEL=CentOS-8-3-2011-x86_64-dvd inst.text inst.cmdline -nographic -monitor none -serial null -display vnc=127.0.0.1:0 -device virtio-serial-pci,id=virtio-serial0 -device virtserialport,bus=virtio-serial0.0,nr=1,chardev=charchannel0,id=channel0,name=org.fedoraproject.anaconda.log.0 -chardev socket,id=charchannel0,host=127.0.0.1,port=40401 -object rng-random,id=virtio-rng0,filename=/dev/random -device virtio-rng-pci,rng=virtio-rng0,id=rng0,bus=pci.0,addr=0x9 2021-05-18 09:24:22,103 DEBUG: Return code: 0 2021-05-18 09:24:22,109 INFO: Running... umount /tmp/lorax.imgutils.nl37yl_1 2021-05-18 09:24:22,127 DEBUG: Return code: 0 2021-05-18 09:24:22,128 INFO: Running... kpartx -v -a -s /tmp/lmc-disk-mr2qmvzx.img 2021-05-18 09:24:22,162 INFO: add map loop0p1 (253:8): 0 8192000 linear 7:0 2048 2021-05-18 09:24:22,163 DEBUG: Return code: 0 2021-05-18 09:24:22,163 INFO: Running... mount /dev/mapper/loop0p1 /tmp/tmpt7b80s77 2021-05-18 09:24:22,171 DEBUG: Return code: 0 2021-05-18 09:24:22,173 INFO: Running... umount /tmp/tmpt7b80s77 2021-05-18 09:24:22,178 DEBUG: Return code: 0 2021-05-18 09:24:22,178 INFO: Running... kpartx -d -s /tmp/lmc-disk-mr2qmvzx.img 2021-05-18 09:24:22,204 DEBUG: Return code: 0 ``` Virt-install log is bigger (will get deleted in 24 hours): https://paste.centos.org/view/a9e864fe Do you have an idea what could be culprit of this? This is clean Fedora 34 installation, I set SELinux to permissive. I cannot find any verbose or debug logging of the tool, I do not understand what lorax is trying to do with the /boot directory (is that a mounted ISO image?). I'd appreciate if you could give my KS a spin to figure out. Or let me know what do I need to test. Thanks.

lzap · May 24, 2021, 7:41am

I was able to build a FDI. I haven’t tested much of its features tho:

Please share if that works for you.