yahoo-eng-team team mailing list archive
-
yahoo-eng-team team
-
Mailing list archive
-
Message #78277
[Bug 1827489] Re: Wrong IPV6 address provided by openstack server create
** Also affects: neutron
Importance: Undecided
Status: New
** Changed in: neutron
Importance: Undecided => High
--
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Compute (nova).
https://bugs.launchpad.net/bugs/1827489
Title:
Wrong IPV6 address provided by openstack server create
Status in neutron:
New
Status in OpenStack Compute (nova):
New
Bug description:
IPV6 address of an interface doesn't have to be derived from its MAC
address. The newer kernels have addr_gen_mode option which controls
the behavior of IPV6 calculation, see
https://www.kernel.org/doc/Documentation/networking/ip-sysctl.txt
I've encountered the problem when I booted up an image (RHEL8 in my
case) which had the addr_gen_mode option set to 1 (means that IPV6
address is randomized) by default. OpenStack (I had Rocky deployment)
didn't recognize this and 'openstack server create' returned wrong
address which lead to tempest failures because thanks to the
'openstack server create' output the tests expected different
addresses on the interfaces.
Steps to reproduce:
$ openstack server create --image <rhel8> --flavor <flavor> --network <network1> --network <network2> --key-name <key_name> instance_name
+-------------------------------------+--------------------------------------------------------------------------------------------------------+
| Field | Value |
+-------------------------------------+--------------------------------------------------------------------------------------------------------+
<output omitted>
| accessIPv4 | |
| accessIPv6 | |
| addresses | tempest-network-smoke--884367252=10.100.0.5; tempest-network-smoke--18828977=2003::f816:3eff:febb:7456 |
<output omitted>
Then ssh to the instance and hit 'ip a' command:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group 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: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1450 qdisc fq_codel state UP group default qlen 1000
link/ether fa:16:3e:48:e8:b5 brd ff:ff:ff:ff:ff:ff
inet 10.100.0.3/28 brd 10.100.0.15 scope global dynamic noprefixroute eth0
valid_lft 86363sec preferred_lft 86363sec
inet6 fe80::f816:3eff:fe48:e8b5/64 scope link
valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1450 qdisc fq_codel state UP group default qlen 1000
link/ether fa:16:3e:bb:74:56 brd ff:ff:ff:ff:ff:ff
inet6 2003::b47f:f400:ecca:2a55/64 scope global dynamic noprefixroute
valid_lft 86385sec preferred_lft 14385sec
inet6 fe80::7615:8d57:775d:fae/64 scope link noprefixroute
valid_lft forever preferred_lft forever
Notice that eth1 interface has an ipv6 address which seems not to be
derived from its mac address. Also notice that the output of
'openstack server create' returned wrong address, a different one than
it's actually set for eth1. It expected that the ipv6 address would be
derived from the mac address but it wasn't.
'openstack server create' should be able to detect the option in the
image and behave accordingly.
To manage notifications about this bug go to:
https://bugs.launchpad.net/neutron/+bug/1827489/+subscriptions
References