MOS 8.0 no ip to instances

[lieni]

Multiple times deployed MOS 8.0
3 Controllers Neutron, VLAN, DVR
1 Compute

The default Test Image Cirros and downloaded Ubuntu Cloudimages dont get a IP after startup.
Windows Cloudbase Image gets an ip adress?!? Strange

On the Windows VM's
I can ping the network Router but no public nets.
I assigne Public IP, internet works.
I cant ping 2. VM in same Subnet.
I cant ping VM from Routernamespace.
i cant curl the metadata IP 169.154.169.254 from Routernamespace.

I can see forwarding to metadata:
ip netns exec qrouter-39628f95-f985-4d92-8f9e-64a1c9f6f613 iptables-save | grep -i redirect
-A neutron-l3-agent-PREROUTING -d 169.254.169.254/32 -i qr-+ -p tcp -m tcp --dport 80 -j REDIRECT --to-ports 8775

On Controller and Compute nodes, Metanet 169.254 has no gateway and no interface in the routing Table is this correct?
route -n
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
0.0.0.0 195.48.9.1 0.0.0.0 UG 0 0 0 br-ex
10.20.0.0 0.0.0.0 255.255.255.0 U 0 0 0 br-fw-admin
169.254.169.254 - 255.255.255.255 !H 0 - 0 -
172.21.1.0 0.0.0.0 255.255.255.0 U 0 0 0 br-storage
192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 br-mgmt
192.168.3.0 0.0.0.0 255.255.255.0 U 0 0 0 br-baremetal
195.48.9.0 0.0.0.0 255.255.255.0 U 0 0 0 br-ex
240.0.0.0 0.0.0.0 255.255.255.252 U 0 0 0 hapr-host
240.0.0.4 0.0.0.0 255.255.255.252 U 0 0 0 vr-host-base

ovs-vsctl show
14fd7c25-e8f5-43b4-bacf-5641ebc18649
    Bridge br-prv
        Port br-prv
            Interface br-prv
                type: internal
        Port phy-br-prv
            Interface phy-br-prv
                type: patch
                options: {peer=int-br-prv}
        Port "p_e52381cd-0"
            Interface "p_e52381cd-0"
                type: internal
    Bridge br-floating
        Port br-floating
            Interface br-floating
                type: internal
        Port phy-br-floating
            Interface phy-br-floating
                type: patch
                options: {peer=int-br-floating}
        Port "p_ff798dba-0"
            Interface "p_ff798dba-0"
                type: internal
    Bridge br-int
        fail_mode: secure
        Port "tap74df66d3-d6"
            tag: 1
            Interface "tap74df66d3-d6"
                type: internal
        Port "tap704a1987-a3"
            tag: 3
            Interface "tap704a1987-a3"
                type: internal
        Port int-br-prv
            Interface int-br-prv
                type: patch
                options: {peer=phy-br-prv}
        Port br-int
            Interface br-int
                type: internal
        Port "qg-46f84cdf-cf"
            tag: 14
            Interface "qg-46f84cdf-cf"
                type: internal
        Port "qr-91bfe3ec-38"
            tag: 1
            Interface "qr-91bfe3ec-38"
                type: internal
        Port "tap81a932d0-1c"
            tag: 2
            Interface "tap81a932d0-1c"
                type: internal
        Port int-br-ironic
            Interface int-br-ironic
                type: patch
                options: {peer=phy-br-ironic}
        Port int-br-floating
            Interface int-br-floating
                type: patch
                options: {peer=phy-br-floating}
    Bridge br-ironic
        Port "p_625d4df9-0"
            Interface "p_625d4df9-0"
                type: internal
        Port br-ironic
            Interface br-ironic
                type: internal
        Port phy-br-ironic
            Interface phy-br-ironic
                type: patch
                options: {peer=int-br-ironic}
    ovs_version: "2.3.1"

Thank you for checking this

[Dmitry Klenov (dklenov)]

Neutron team, please check if issue is located on Neutron side. It might be the case that Fuel creates incorrect L2 topology. Please investigate.

[Oleg Bondarev (obondarev)]

I doubt it's a neutron issue since 8.0 should be regularly tested by out CI and the issue of a test VM not getting an IP or not getting connectivity (a very basic test!) should be instantly caught. So I suspect it's some configuration/deployment issue.

Please attach a diagnostic snapshot for further investigation. Please also describe your virtual network topology: networks, routers, VMs (with IDs) and the output of 'ovs-vsctl show'/'brctl show' from all controllers and compute nodes.

[lieni (oliver-lienhard-b)]

I seems that the connection from Qrouter on Compute nodes to controller nodes dosent work with the VLAN deploy mode.
We checked same settings with VXLAN, that works with no problems. Only no bare metal available :-(

[lieni (oliver-lienhard-b)]

Same error on 7.0 inkl. Updates.

[Timur Nurlygayanov (tnurlygayanov)]

Hi lieni, could you please attach diagnostic snapshot here? It will help to understand the root of the issue.

Thank you!

[lieni (oliver-lienhard-b)]

Sorry this deployment was allready deleted.
We have deployed vxlan mode, that works fine.
with vlans severale times failed.

[Alexander Ignatov (aignatov)]

Moving it to invalid state since it's not reproducible, diagnostic snapshot was not attached and seems will not be added in future.