CentOS 8 Installer Bug

Home » CentOS » CentOS 8 Installer Bug
CentOS 2 Comments

CentOS 8.1.1911 has a bug in the installer. It is not possible to use more than one NFS repo via kickstart repo directive as the installer tries to mount all of them on /run/install/.nfs. You get this

DBG payload: /run/install/.nfs already has something mounted on it

message for all but the first one. The problem was reported for Fedora 29/30.

https://bugzilla.redhat.com/show_bug.cgi?id69711

Do we know if this is fixed in RHEL 8.2?

2 thoughts on - CentOS 8 Installer Bug

  • Has anyone managed to create an encrypted disk partition with CentOS 8 kickstart?

    1 reqpart –add-boot
    2 part /boot –fstype ext3 –size24 –ondrive=sda
    3 part pv.1 –size=1 –grow –ondrive=sda –encrypted –pashphrase=”bla”
    4 volgroup vol0 pv.100000
    5 logvol / –vgname=vg_00 –name=lv_root –size2400 –fstype=ext4
    …

    No matter what I specify in line 3, the installer bombs out. The only way to successful kickstart is to not encrypt the lvm pv. The contents of the passphrase are not to blame, installer still dies after accepting a passphrase from the console.

    15:57:45,280 DBG kickstart: Looking for platform-specific boot requirements.
    15:57:45,281 DBG kickstart: Applying requirements:
    PartSpec instance (0x7fbd40f44320)

  • isdtor writes:

    So, yeah, apologies for the typos. The actual ks file is correct and consistent.

    The problem was line 3 – apparently, “–size=1 –grow” doesn’t cut it anymore. Increasing size to a reasonable number, e.g in the order of “sum of all lvols”, got me over the hump.