vxlan ttl defaults to 1
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack-Ansible |
Fix Released
|
Medium
|
Jonathan Rosser |
Bug Description
No value is set for ttl in the [vxlan] section of linuxbridge_
Here is an example for an instance arp packet leaving a compute host:
18:33:34.303003 7c:fe:90:de:08:20 > 01:00:5e:01:01:01, ethertype 802.1Q (0x8100), length 96: vlan 1941, p 0, ethertype IPv4, (tos 0x0, ttl 1, id 61202, offset 0, flags [none], proto UDP (17), length 78)
10.
fa:16:3e:22:7a:77 > ff:ff:ff:ff:ff:ff, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 192.168.100.17 tell 192.168.100.12, length 28
The outer ethernet frame ttl=1 means that the vxlan packets are dropped at the next hop router in a l3 environment as the ttl is decremented to 0 at that point.
This has no impact in an AIO, MNAIO or l2 underlay deployment, but will break with a l3 underlay as described in the documentation [1].
Two things could be done:
* By default set a more appropriate ttl in the linuxbridge ini file.
* Include the necessary variables in the deploy guide example [1].
neutron_
vxlan:
ttl: 32
[1] https:/
Changed in openstack-ansible: | |
status: | New → Confirmed |
importance: | Undecided → Medium |
assignee: | nobody → Jean-Philippe Evrard (jean-philippe-evrard) |
good catch. Seems legit to me.