metadata route is absent and metadata server unavailable for instance with network w/o gateway
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Compute (nova) |
Expired
|
Undecided
|
Unassigned | ||
neutron |
Invalid
|
Undecided
|
Unassigned |
Bug Description
ubuntu 14.04
latest devstack (28.08.2016)
enabled services: nova, glance, cinder, keystone, horizon, neutron, neutron-vpnaas, ec2-api
steps to reproduce:
0) source demo credentials
1) create network
2) create subnet without gateway
3) boot instance from cirros image
4) run vnc console
5) run 'route -n':
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
10.10.0.0 10.10.2.1 255.255.0.0 UG 0 0 0 eth0
10.10.2.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
6) run 'curl http://
curl: (7) Failed to connect to 169.254.169.254: Network is unreachable
if user runs instance with predefined network that all works well:
$ route -n
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
0.0.0.0 10.10.1.1 0.0.0.0 UG 0 0 0 eth0
10.10.0.0 10.10.1.1 255.255.0.0 UG 0 0 0 eth0
10.10.1.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
169.254.169.254 10.10.1.1 255.255.255.255 UGH 0 0 0 eth0
according to gating jobs it was broken between "Aug 25 2:05 PM" (last check pipeline - https:/
and "Aug 26 1:05 AM" (patch 4 first check pipeline - https:/
tags: | added: api metadata |
tags: | added: newton-rc-potential |
Changed in nova: | |
status: | New → Incomplete |
This was when we changed the default to use metadata server all the time. It sounds like previously the ec2api didn't work if it was using metadata server instead of config drive.
Is there some default additional metadata route that needs to be put into neutron during such a config?