David Lehman wrote:
On Thu, 2009-04-09 at 15:52 +0200, Radek Vykydal wrote:
@@ -881,7 +882,11 @@ class VolGroup(commands.volgroup.FC3_VolGroup):
if not device:
raise KicsktartValueError, formatErrorMsg(self.lineno, msg="Specified nonexistent VG %s in volgroup command" % vgd.vgname)
- devicetree.registerAction(ActionCreateFormat(device))
+ for lv in device.lvs:
+ format = getFormat(lv.format.type,
+ mountpoint=lv.format.mountpoint,
+ mountopts=lv.format.mountopts)
+ devicetree.registerAction(ActionCreateFormat(lv, format))
This doesn't make any sense to me. I do agree with removing the format
create action for the VG, since the VG has no formatting. However, those
LVs will be formatted according to the logvol commands -- why do this
magic in the volgroup handler?
I was trying to make sense of volgroup option --useexisting:
"Use an existing volume group and reformat it."
What I came up to is that the intetntion of the option could
have been to use it when only specifying volgroup without any following
commands for logical volumes, e.g.:
volgroup vg_dhcp69 --useexisting
would reformat all logical volumes of the volume group.
I don't see it as a useful case, but if we don't want
to support it, then --useexisting seems redundant to --noformat
for volgroup, we should probably remove it from documentation or correct it,
and do all format/noformat specification of existing lvs only
in logvol command e.g.:
volgroup vg_dhcp69 --noformat
logvol /data --name=LogVol00 --vgname=vg_dhcp69 --fstype=ext3 --useexisting
which would format and use existing lv,
or
volgroup vg_dhcp69 --noformat
logvol /data --name=LogVol00 --vgname=vg_dhcp69 --noformat
which would use existing lv.
I am ok with only removing format action of volume group as you
suggested. Should I push with this change?
Radek
_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
