Update kickstarts to remove cgts-vg as needed

In some scenarious, such as when the cgts-vg volume group spans
multiple disks, the disk wiping in the kickstarts is insufficient to
clear existing LVM metadata, and anaconda reports a failure to create
the cgts-vg volume group as a result.

This update adds explicit cleanup of the logical and physical volumes
associated with cgts-vg, if not already wiped by the user prior to
installation, and removes cgts-vg itself. This is done ahead of the
anaconda partitioning, avoiding any conflict with automatic creation
of the new cgts-vg.

Change-Id: Id33cab02e1cecbfee8035cc55213c99ef3cea34d
Closes-Bug: 1854082
Signed-off-by: Don Penney <don.penney@windriver.com>
This commit is contained in:
Don Penney 2019-11-26 11:52:49 -05:00
parent 97d07e0f30
commit eb851a9da0

View File

@ -66,6 +66,11 @@ fi
# Deactivate existing volume groups to avoid Anaconda issues with pre-existing groups # Deactivate existing volume groups to avoid Anaconda issues with pre-existing groups
vgs --noheadings -o vg_name | xargs --no-run-if-empty -n 1 vgchange -an vgs --noheadings -o vg_name | xargs --no-run-if-empty -n 1 vgchange -an
# Remove volumes and group for cgts-vg, if any
lvremove --force cgts-vg
pvs --select 'vg_name=cgts-vg' --noheadings -o pv_name | xargs --no-run-if-empty pvremove --force --force --yes
vgs --select 'vg_name=cgts-vg' --noheadings -o vg_name | xargs --no-run-if-empty vgremove --force
ONLYUSE_HDD="" ONLYUSE_HDD=""
if [ "$(curl -sf http://pxecontroller:6385/v1/upgrade/$(hostname)/in_upgrade 2>/dev/null)" = "true" ]; then if [ "$(curl -sf http://pxecontroller:6385/v1/upgrade/$(hostname)/in_upgrade 2>/dev/null)" = "true" ]; then
# In an upgrade, only wipe the disk with the rootfs and boot partition # In an upgrade, only wipe the disk with the rootfs and boot partition