Browse Source

Handle edges with different number of tunnels

Edge appliances with different number of tunnels per vnic might exit
within the system.
That could happen due to a change in the config file after the system
has been running for a while and edge appliances already exist.
The router interface allocation logic should support this edge case.

Change-Id: I47b72072a44ad40225714295aabcc5b7198eb71f
changes/54/704754/2
Kobi Samoray 2 years ago
committed by asarfaty
parent
commit
76b2648336
  1. 9
      vmware_nsx/db/nsxv_db.py

9
vmware_nsx/db/nsxv_db.py

@ -212,11 +212,18 @@ def clean_edge_vnic_binding(session, edge_id):
def allocate_edge_vnic(session, edge_id, network_id):
"""Allocate an available edge vnic to network."""
# get vnic count of specific edge
bindings = (session.query(nsxv_models.NsxvEdgeVnicBinding).
filter_by(edge_id=edge_id,
vnic_index=1).all())
vnic_tunnels_per_index = len(bindings)
with session.begin(subtransactions=True):
bindings = (session.query(nsxv_models.NsxvEdgeVnicBinding).
filter_by(edge_id=edge_id, network_id=None).all())
for binding in bindings:
if binding['tunnel_index'] % constants.MAX_TUNNEL_NUM == 1:
if binding['tunnel_index'] % vnic_tunnels_per_index == 1:
binding['network_id'] = network_id
session.add(binding)
return binding

Loading…
Cancel
Save