Ask Your Question

Instance failling with VLAN network but not with VXLAN

asked 2019-06-10 10:19:07 -0500

VictorCM gravatar image

updated 2019-06-12 10:50:23 -0500

Hi, I am working on an isolated tenant with no connection to internet and testing with differents configurations and when I create an instance attached to a VXLAN network everything goes as expected, but when I attach a VLAN to the instance it takes too long to spawn up the VM and it ignores the config-init I use..

Probably is due to the lack of knowledge I have on this area but I dont understand why this is happening, can anyone help me to understand this? I suppose it is because how these types of networks work on OpenStack

Info about topology and networks: (image)

Thank you so much

PD: The VM (with VLAN net) stuck on this line ~10 min ((screenshot console)) and its log is:

[    6.062736] iscsi: registered transport (iser)
[[0;32m  OK  [0m] Started Load Kernel Modules.
     Mounting FUSE Control File System...
     Starting Apply Kernel Variables...
[[0;32m  OK  [0m] Mounted FUSE Control File System.
[[0;32m  OK  [0m] Started Flush Journal to Persistent Storage.
[[0;32m  OK  [0m] Started udev Coldplug all Devices.
[[0;32m  OK  [0m] Started Apply Kernel Variables.
[[0;32m  OK  [0m] Started Create Static Device Nodes in /dev.
     Starting udev Kernel Device Manager...
[[0;32m  OK  [0m] Started Monitoring of LVM2 mirrors, dmeventd or progress polling.
[[0;32m  OK  [0m] Reached target Local File Systems (Pre).
[[0;32m  OK  [0m] Reached target Local File Systems.
     Starting Initial cloud-init job (pre-networking)...
     Starting Create Volatile Files and Directories...
     Starting Set console font and keymap...
     Starting Tell Plymouth To Write Out Runtime Data...
     Starting Commit a transient machine-id on disk...
     Starting LSB: AppArmor initialization...
[[0;32m  OK  [0m] Started Create Volatile Files and Directories.
     Starting Network Time Synchronization...
     Starting Update UTMP about System Boot/Shutdown...
[[0;32m  OK  [0m] Started Update UTMP about System Boot/Shutdown.
[[0;32m  OK  [0m] Started Tell Plymouth To Write Out Runtime Data.
[[0;32m  OK  [0m] Started Network Time Synchronization.
[[0;32m  OK  [0m] Started Commit a transient machine-id on disk.
[[0;32m  OK  [0m] Reached target System Time Synchronized.
[[0;32m  OK  [0m] Started udev Kernel Device Manager.
[[0;32m  OK  [0m] Started Dispatch Password Requests to Console Directory Watch.
[[0;32m  OK  [0m] Found device /dev/ttyS0.
[[0;32m  OK  [0m] Listening on Load/Save RF Kill Switch Status /dev/rfkill Watch.
[[0;32m  OK  [0m] Started LSB: AppArmor initialization.
[[0;32m  OK  [0m] Started Set console font and keymap.
[[0;32m  OK  [0m] Created slice system-getty.slice.

UPDATE First response from Bernd Bausch:


type_drivers = flat,vlan,vxlan
tenant_network_types = vxlan
mechanism_drivers = openvswitch,l2population
extension_drivers = port_security

network_vlan_ranges = physnet1
flat_networks = physnet1
bridge_mappings = physnet1:br-ex
edit retag flag offensive close merge delete


I would check for problems with the mapping of VLAN to the network infrastructure under the cloud. Assuming you use ML2, this is done in the mlw-conf.ini file and the Neutron agent’s config file.

You might also find relevant messages in Neutron’s log files.

Bernd Bausch gravatar imageBernd Bausch ( 2019-06-11 00:35:48 -0500 )edit

Hi, thanks for your comment. I have been these days redeploying the environment to check that it wasnt an error of that part and checking logs.. There is no error on logs so I wrote again the ml_conf.ini and neutron.conf. I am updating the post with this info so we can check if there is any error..

VictorCM gravatar imageVictorCM ( 2019-06-12 10:46:35 -0500 )edit

I think you have to provide one or more ranges like

network_vlan_ranges = physnet1:1100:1110

where 1100 and 1110 are VLAN IDs.

Bernd Bausch gravatar imageBernd Bausch ( 2019-06-12 17:15:53 -0500 )edit

Sorry I have been busy at work and I couldnt check before. I did it but issue didnt change. With cirros this is not happening because it is not loading a cloud-config. Is there any known issue related to this?

VictorCM gravatar imageVictorCM ( 2019-06-25 09:19:07 -0500 )edit

1 answer

Sort by » oldest newest most voted

answered 2019-06-13 02:43:31 -0500

Moss gravatar image

I would use cirros image to get some more verbose console logs from within booting vm.

I'm using vlan&vxlan tenant networks - check and compare my working configs below (openstack pike release):


egrep -v '#|^$'  /etc/neutron/plugins/ml2/ml2_conf.ini
type_drivers = vlan,vxlan
tenant_network_types = vlan,vxlan
mechanism_drivers = openvswitch,l2population
extension_drivers = port_security,dns_domain_ports
network_vlan_ranges = vlan:300:500
vni_ranges = 1:1000
enable_ipset = True


egrep -v '#|^$' /etc/neutron/plugins/ml2/openvswitch_agent.ini
tunnel_types = vxlan
l2_population = true
arp_responder = true
bridge_mappings = vlan:br-vlan
local_ip = x.x.x.2
firewall_driver = neutron.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver

It should help you to pinpoint missing configuration on your side.

edit flag offensive delete link more


Hi! Thanks for your responce. Sorry to answer back now. I have check my ml2_conf,ini and its pretty equal. I have test the deployment with instances with cirros os and both spawn good (attached to VLAN one a bit slow, but good enough) but they didnt load the cloud-config file I was using

VictorCM gravatar imageVictorCM ( 2019-06-25 09:51:39 -0500 )edit

( I used the same cloud-config as in ubuntu instance, so probably that is the problem). So, by now, I supose that the problem related to the VLAN can be: - Instances attached to a VLAN network need a config that im not using - Related to cloud-config?

VictorCM gravatar imageVictorCM ( 2019-06-25 09:57:29 -0500 )edit

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2019-06-10 10:19:07 -0500

Seen: 231 times

Last updated: Jun 13 '19