bionic: LXD containers don't get fan network IP addresses
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Canonical Juju |
Fix Released
|
High
|
Unassigned | ||
2.3 |
Won't Fix
|
High
|
Unassigned | ||
cloud-init (Ubuntu) |
Invalid
|
Undecided
|
Unassigned | ||
lxd (Ubuntu) |
Invalid
|
Undecided
|
Unassigned |
Bug Description
Juju Version: 2.3.4
Substrate: OpenStack
I'm deploying a bundle which makes use of LXD containers, leveraging fan networking for communication between containers and hosts; containers start and have the following config snippet:
devices:
eth0:
hwaddr: 00:16:3e:01:7e:0f
mtu: "8908"
name: eth0
nictype: bridged
parent: fan-252
type: nic
The container has a NIC:
root@juju-
1: lo: <LOOPBACK,
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
13: eth0@if14: <BROADCAST,
link/ether 00:16:3e:01:7e:0f brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet6 fe80::216:
valid_lft forever preferred_lft forever
which is marked up and the host devices:
# ip addr
1: lo: <LOOPBACK,
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: ens3: <BROADCAST,
link/ether fa:16:3e:c6:8a:fa brd ff:ff:ff:ff:ff:ff
inet 10.5.0.4/16 brd 10.5.255.255 scope global dynamic ens3
valid_lft 85454sec preferred_lft 85454sec
inet6 fe80::f816:
valid_lft forever preferred_lft forever
3: fan-252: <BROADCAST,
link/ether e6:86:08:94:73:dc brd ff:ff:ff:ff:ff:ff
inet 252.0.4.1/8 scope global fan-252
valid_lft forever preferred_lft forever
inet6 fe80::e486:
valid_lft forever preferred_lft forever
4: ftun0: <BROADCAST,
link/ether e6:86:08:94:73:dc brd ff:ff:ff:ff:ff:ff
inet6 fe80::e486:
valid_lft forever preferred_lft forever
5: ovs-system: <BROADCAST,
link/ether 82:2c:33:fd:91:81 brd ff:ff:ff:ff:ff:ff
6: br-int: <BROADCAST,
link/ether 2e:5f:c2:1e:4a:40 brd ff:ff:ff:ff:ff:ff
7: lxdbr0: <BROADCAST,
link/ether 42:f2:f3:0e:18:37 brd ff:ff:ff:ff:ff:ff
inet 10.201.15.1/24 scope global lxdbr0
valid_lft forever preferred_lft forever
inet6 fe80::40f2:
valid_lft forever preferred_lft forever
8: br-ex: <BROADCAST,
link/ether 62:c9:e1:f9:44:48 brd ff:ff:ff:ff:ff:ff
10: veth9PCNPR@if9: <BROADCAST,
link/ether fe:54:70:62:e2:a5 brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet6 fe80::fc54:
valid_lft forever preferred_lft forever
12: vethH72G7D@if11: <BROADCAST,
link/ether fe:38:7d:fc:da:90 brd ff:ff:ff:ff:ff:ff link-netnsid 1
inet6 fe80::fc38:
valid_lft forever preferred_lft forever
14: vethRVTEFS@if13: <BROADCAST,
link/ether fe:7a:0d:78:95:71 brd ff:ff:ff:ff:ff:ff link-netnsid 2
inet6 fe80::fc7a:
valid_lft forever preferred_lft forever
look to be mapped correctly into the fan network - the fan dnsmasq process is running; and then I found this in the cloud-init log:
2018-03-15 10:35:03,625 - stages.py[WARNING]: Failed to rename devices: Failed to apply network config names. Found bad network config version: None
2018-03-15 10:35:03,627 - util.py[WARNING]: failed stage init-local
failed run of stage init-local
-------
Traceback (most recent call last):
File "/usr/lib/
ret = functor(name, args)
File "/usr/lib/
init.
File "/usr/lib/
return self.distro.
File "/usr/lib/
dev_names = self._write_
File "/usr/lib/
return self._supported
File "/usr/lib/
renderer.
File "/usr/lib/
network_
File "/usr/lib/
content = self._render_
File "/usr/lib/
if network_
AttributeError: 'NoneType' object has no attribute 'version'
ProblemType: Bug
DistroRelease: Ubuntu 18.04
Package: lxd 3.0.0~beta3-
ProcVersionSign
Uname: Linux 4.15.0-10-generic x86_64
ApportVersion: 2.20.8-0ubuntu10
Architecture: amd64
Date: Thu Mar 15 10:42:58 2018
Ec2AMI: ami-00000456
Ec2AMIManifest: FIXME
Ec2Availability
Ec2InstanceType: m1.xlarge
Ec2Kernel: unavailable
Ec2Ramdisk: unavailable
JournalErrors:
-- Logs begin at Thu 2018-03-15 10:29:35 UTC, end at Thu 2018-03-15 10:43:02 UTC. --
Mar 15 10:29:59 hostname iscsid[1048]: iSCSI daemon with pid=1051 started!
Mar 15 10:32:46 hostname systemd[1]: Failed to start Router advertisement daemon for IPv6.
ProcEnviron:
TERM=screen
PATH=(custom, no user)
LANG=C.UTF-8
SHELL=/bin/bash
SourcePackage: lxd
UpgradeStatus: No upgrade log present (probably fresh install)
summary: |
- LXD containers don't get fan network IP addresses + bionic: LXD containers don't get fan network IP addresses |
tags: | added: uosci |
Changed in juju: | |
status: | New → Triaged |
importance: | Undecided → High |
milestone: | none → 2.4-beta1 |
Changed in juju: | |
milestone: | 2.4-beta1 → none |
Changed in juju: | |
milestone: | none → 2.5-beta1 |
Changed in juju: | |
assignee: | nobody → Richard Harding (rharding) |
Changed in juju: | |
milestone: | 2.5-beta1 → 2.5-beta2 |
Changed in juju: | |
assignee: | Richard Harding (rharding) → nobody |
milestone: | 2.5-beta2 → 2.5.1 |
Changed in juju: | |
milestone: | 2.5.1 → 2.5.2 |
I'll note that I tried just launching a bionic container using snap
3.0.0.beta5 and after "lxc launch ubuntu:x" and "lxc launch ubuntu-daily:b"
neither of them came up with an IP address.
I might have broken my networking on this machine because I was trying to
install stock Juju which tries to bring in stock lxd, ignoring the snap,
but I'm also seeing failures for containers to get IP addresses.
On Thu, Mar 15, 2018 at 2:47 PM, James Page <email address hidden> wrote:
> Public bug reported: 4e2443- 0-lxd-0: ~# ip addr UP,LOWER_ UP> mtu 65536 qdisc noqueue state UNKNOWN group MULTICAST, UP,LOWER_ UP> mtu 8908 qdisc noqueue 3eff:fe01: 7e0f/64 scope link UP,LOWER_ UP> mtu 65536 qdisc noqueue state UNKNOWN group MULTICAST, UP,LOWER_ UP> mtu 8958 qdisc fq_codel state 3eff:fec6: 8afa/64 scope link MULTICAST, UP,LOWER_ UP> mtu 8908 qdisc noqueue 8ff:fe94: 73dc/64 scope link MULTICAST, UP,LOWER_ UP> mtu 8908 qdisc noqueue master 8ff:fe94: 73dc/64 scope link MULTICAST> mtu 1500 qdisc noop state DOWN group MULTICAST> mtu 1500 qdisc noop state DOWN group
>
> Juju Version: 2.3.4
> Substrate: OpenStack
>
> I'm deploying a bundle which makes use of LXD containers, leveraging fan
> networking for communication between containers and hosts; containers
> start and have the following config snippet:
>
> devices:
> eth0:
> hwaddr: 00:16:3e:01:7e:0f
> mtu: "8908"
> name: eth0
> nictype: bridged
> parent: fan-252
> type: nic
>
>
> The container has a NIC:
>
> root@juju-
> 1: lo: <LOOPBACK,
> default qlen 1000
> link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> inet 127.0.0.1/8 scope host lo
> valid_lft forever preferred_lft forever
> inet6 ::1/128 scope host
> valid_lft forever preferred_lft forever
> 13: eth0@if14: <BROADCAST,
> state UP group default qlen 1000
> link/ether 00:16:3e:01:7e:0f brd ff:ff:ff:ff:ff:ff link-netnsid 0
> inet6 fe80::216:
> valid_lft forever preferred_lft forever
>
> which is marked up and the host devices:
>
> # ip addr
> 1: lo: <LOOPBACK,
> default qlen 1000
> link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> inet 127.0.0.1/8 scope host lo
> valid_lft forever preferred_lft forever
> inet6 ::1/128 scope host
> valid_lft forever preferred_lft forever
> 2: ens3: <BROADCAST,
> UP group default qlen 1000
> link/ether fa:16:3e:c6:8a:fa brd ff:ff:ff:ff:ff:ff
> inet 10.5.0.4/16 brd 10.5.255.255 scope global dynamic ens3
> valid_lft 85454sec preferred_lft 85454sec
> inet6 fe80::f816:
> valid_lft forever preferred_lft forever
> 3: fan-252: <BROADCAST,
> state UP group default qlen 1000
> link/ether e6:86:08:94:73:dc brd ff:ff:ff:ff:ff:ff
> inet 252.0.4.1/8 scope global fan-252
> valid_lft forever preferred_lft forever
> inet6 fe80::e486:
> valid_lft forever preferred_lft forever
> 4: ftun0: <BROADCAST,
> fan-252 state UNKNOWN group default qlen 1000
> link/ether e6:86:08:94:73:dc brd ff:ff:ff:ff:ff:ff
> inet6 fe80::e486:
> valid_lft forever preferred_lft forever
> 5: ovs-system: <BROADCAST,
> default qlen 1000
> link/ether 82:2c:33:fd:91:81 brd ff:ff:ff:ff:ff:ff
> 6: br-int: <BROADCAST,
> defau...