=== cpaelzer__ is now known as cpaelzer | ||
sdhd-sascha | hi, | 12:52 |
---|---|---|
sdhd-sascha | I have a bond-device and added it to a bridge. | 12:52 |
sdhd-sascha | I want to use this bridge for lxd and kvm, to expose all machines into the physical network. | 12:52 |
sdhd-sascha | Should i enable stp ? Because, if i add `stp: yes` to netplan - then my network is not reachable. | 12:52 |
sdhd-sascha | (Not sure how to debug this issue. My network skills are not the best ) | 12:52 |
sdhd-sascha | Currently it's ubuntu 20.04 - but had the same effect before in 19.10 | 12:52 |
sdhd-sascha | Should i give the bond0 all mac-addresses from all physical interfaces ? (Currently don't know how to look into the arp-table of my switch) | 13:01 |
sdhd-sascha | E. g. MAAS gives the bond only one mac-address from one device (not sure, if the bond-mode does control this) | 13:02 |
nils_ | sdhd-sascha, not sure if this is strictly kernel related. | 13:04 |
sdhd-sascha | nils_: i'm too. Whom could i ask ? | 13:05 |
sdhd-sascha | Some Ceph or Openstack people ... I look, for other channels ;-) thank you | 13:06 |
nils_ | the bond device is connected to a switch? | 13:09 |
nils_ | It may be that enabling STP on the bridge interferes with the switch. | 13:09 |
sdhd-sascha | nils_: yes. Good hint. :-) | 13:21 |
sdhd-sascha | I found #openstack. And there i get the information, that no `stp` is needed, because i didn't create loops on the host | 13:21 |
sdhd-sascha | Thank you :-) | 13:21 |
* sdhd-sascha really need to refresh my network knowledge | 13:22 | |
=== mhcerri_ is now known as mhcerri | ||
zx2c4 | apw: any repo where i can follow along your dkms reworking? ive been poking around launchpad to get a handle on it but havent found much | 15:32 |
apw | zx2c4, oh no, that is all local right now; the fight i have been having with the librarian and provides is over | 15:39 |
zx2c4 | apw: oh my... | 15:53 |
zx2c4 | O_o | 15:53 |
zx2c4 | do you mean this whole thing with linux-module-wireguard "providing" wireguard-modules, which the wireguard metapackage depends on? | 15:54 |
apw | zx2c4, the support that would be used for that; but for the existing builds, nvidia and hte like | 16:29 |
zx2c4 | ah gotcha | 16:31 |
apw | but the existing stuff was a bit of a mess to add to, and bugs in the first phase take hours to build | 16:33 |
zx2c4 | zoinks | 16:33 |
zx2c4 | well hopefully wireguard will just be "drop in" | 16:33 |
zx2c4 | should have a relatively sane build system compared to nvidia.... | 16:33 |
apw | yes, it should be easy, nvidia not so much | 16:47 |
zx2c4 | apw: well let me know when i can try it out / when youve got something publish pushed | 16:57 |
fling | Is eoan at 5.4 now? | 20:52 |
fling | Am I supposed to get shiftfs patches from eoan if I want them for 5.4? | 20:52 |
fling | probably not | 21:03 |
gpiccoli | fling, eoan is 5.3 - the new release (devel), called Focal, is 5.4 AFAIK | 21:04 |
fling | gpiccoli: how do I clone it's repo using linus tree as base? | 21:07 |
fling | git clone --reference linux focal.git | 21:08 |
fling | what is the focal git repo? :P | 21:08 |
sarnold | fling: https://kernel.ubuntu.com/git/ubuntu/ looks to have one | 21:09 |
fling | thanks | 21:09 |
gpiccoli | I've been using this one fling : git://kernel.ubuntu.com/kernel-ppa/mirror/ubuntu-focal.git | 21:09 |
fling | n=5401 ; for c in $(git log fs/shiftfs.c | grep ^commit | sed s/commit\ // | tac) ; do git format-patch -1 --start-number $n $c ; n=$(($n+1)) ; done | 21:12 |
fling | this is how I get shiftfs patches ^ is there a better way to filter out what I need? | 21:13 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!