Openstack HA : internal vrouter id and external vrouter id in keepalived conf are configured same

[venu kolli]

internal vrouter id and external vrouter id in keepalived conf are configured same if those parameters were not specified in ha.env configuration.

default values for internal and external are 100 and 200 but now both were defaulted to 100

   'internal_virtual_router_id':
   'external_virtual_router_id':

[venu kolli (vkolli)]

Workaround :

When HA failures are being performed , please make sure following changes were done.

stop keepalived service

modify the keepalived configuration /etc/keepalived/keepalived.conf of internal vrouter id to be 100 and external vrouter id to 200 .

start keepalived service

[Jeba Paulaiyan (jebap)]

Workaround for release-notes:

When a cluster is provisioned in openstack HA mode, if the user does not explicitly specify the values for the following parameters, then the system assumes 100 for these values. Having same values for both of these parameters is not recommended as it might exhibit issues when a node goes down and another nodes takes over the vIP.

 'internal_virtual_router_id' : 180, #Default = 100
 'external_virtual_router_id' : 190, #Default = 100

It is highly recommended that the user specify the values for the above parameters in testbed.py while provisioning.

On an already provisioned system, the user can edit the values in /etc/keepalived/keepalived.conf file. Look for virtual_router_id in the file. Note: After editing the file, restart keepalived process by "service keepalived stop" and "service keepalived start".

[Abhay Joshi (abhayj)]

This is already fixed (at least with SM provisioning - Please see bug 1379559 and https://review.opencontrail.org/#/c/13350/, https://review.opencontrail.org/#/c/13351/.