r1981 - Don't hardcode net.ifnames=0 since KARGS defaults to including now in r1977

26 views
Skip to first unread message

com...@warewulf.lbl.gov

unread,
Apr 9, 2016, 10:56:46 PM4/9/16
to warewul...@lbl.gov
Author: bsallen
Date: 2016-04-09 19:56:44 -0700 (Sat, 09 Apr 2016)
New Revision: 1981

Modified:
trunk/provision/lib/Warewulf/Provision/Pxelinux.pm
Log:
Don't hardcode net.ifnames=0 since KARGS defaults to including now in r1977

Modified: trunk/provision/lib/Warewulf/Provision/Pxelinux.pm
===================================================================
--- trunk/provision/lib/Warewulf/Provision/Pxelinux.pm 2016-04-08 21:27:28 UTC (rev 1980)
+++ trunk/provision/lib/Warewulf/Provision/Pxelinux.pm 2016-04-10 02:56:44 UTC (rev 1981)
@@ -254,9 +254,6 @@
if ($console) {
print PXELINUX "console=tty0 console=$console ";
}
- if (! $ifrename) {
- print PXELINUX "net.ifnames=0 ";
- }
if (scalar(@masters) > 0) {
my $master = join(",", @masters);
print PXELINUX "wwmaster=$master ";

Allen, Benjamin S.

unread,
May 29, 2016, 2:32:26 PM5/29/16
to warewul...@lbl.gov
Should we pull the ifrename code out that we added r1921 since we're using KARGS to set net.ifnames=0 biosdevname=0 in r1983 and r1977 by default?

Currently the CLI doesn't appear to work for --ifrename:

$ wwsh node set --ifrename=true n01
Unknown option: ifrename

Lastly, with the changed added in r1988 and r1989 should we actually default to disabling net.ifnames and biosdevnames?

Ben
> --
> You received this message because you are subscribed to the Google Groups "Warewulf-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to warewulf-deve...@lbl.gov.
>

Jason Stover

unread,
May 29, 2016, 2:43:49 PM5/29/16
to warewul...@lbl.gov
ifrename is under provision not node ... though I'm not sure if it's
actually used anywhere. :/ There's only this in Pxelinux.pm:

my $ifrename = $nodeobj->ifrename();

But nothing done with that '$ifrename' value.

And I would prefer having that as the default still. If KARGS are
actually set with something, then those should be removed... so if it
specifically gets set to just... 'quiet' or 'debug' or something, then
those defaults shouldn't apply.

But that's my personal choice as I... dislike the new naming and would
really like to ask whoever came up with it what they were on at the
time. ;)

-J

Allen, Benjamin S.

unread,
May 29, 2016, 3:14:48 PM5/29/16
to warewul...@lbl.gov
Oh doh... on node versus provision.

Yea this r1981 commit removed the effect of --ifrename. Greg you added --ifrename to provision, what should we do with it?

Also should we make the KARGS default configurable in /etc/warewulf/default/provision.conf or similar?

Ben

Jason Stover

unread,
May 29, 2016, 3:28:29 PM5/29/16
to warewul...@lbl.gov
I say if the effects were removed, get rid of the other code since
it's now useless. I can't actually do anything I wanted to this
weekend, but I can maybe tomorrow night start digging in.

I suppose having configurable default would be okay. Would you say
having a default value if one doesn't exist, or something that's
always appended to the kargs value be better? Meaning, so there's
always 'debug' instead of quiet, etc...

-J


On Sun, May 29, 2016 at 2:14 PM, Allen, Benjamin S.

Allen, Benjamin S.

unread,
May 29, 2016, 6:22:54 PM5/29/16
to warewul...@lbl.gov
Something like the existing values in /etc/warewulf/default/provision.conf, where the config option will be used as the default on node creation.

Ben

Jason Stover

unread,
Jun 3, 2016, 7:00:31 PM6/3/16
to warewul...@lbl.gov
Hi Ben,

In case you miss it I committed r1991 which has the KARGS change in it.

-J


On Sun, May 29, 2016 at 5:22 PM, Allen, Benjamin S.
Reply all
Reply to author
Forward
0 new messages