This required a trivial refactor of two existing cases in the codebase.
These two cases were perfectly correct, but the check uncovered a 3rd
case which was a real bug (fixed separately). The new versions also
make it clear that if the loop fails to break early then the 'result' is
None (and thus an error) and not simply the last element. On balance,
it's probably worth enforcing this small inconvenience to coding style.
Change-Id: I780a95241f1454c6886d91f980eb9ada7678a119
Related-Bug: #
1362466
signature-differs,
star-args,
super-init-not-called,
- undefined-loop-variable,
unnecessary-lambda,
unnecessary-pass,
unpacking-non-sequence,
'command': 'bridge fdb',
'mode': 'VXLAN UCAST'})
return False
- for segmentation_id in moves.xrange(1, constants.MAX_VXLAN_VNI + 1):
+
+ test_iface = None
+ for seg_id in moves.xrange(1, constants.MAX_VXLAN_VNI + 1):
if not ip_lib.device_exists(
- self.get_vxlan_device_name(segmentation_id)):
+ self.get_vxlan_device_name(seg_id)):
+ test_iface = self.ensure_vxlan(seg_id)
break
else:
LOG.error(_LE('No valid Segmentation ID to perform UCAST test.'))
return False
- test_iface = self.ensure_vxlan(segmentation_id)
try:
utils.execute(
cmd=['bridge', 'fdb', 'append', constants.FLOODING_ENTRY[0],
def _get_server_health_option(config):
"""return the first active health option."""
- for monitor in config['healthmonitors']:
+ for m in config['healthmonitors']:
# not checking the status of healthmonitor for two reasons:
# 1) status field is absent in HealthMonitor model
# 2) only active HealthMonitors are fetched with
# LoadBalancerCallbacks.get_logical_device
- if monitor['admin_state_up']:
+ if m['admin_state_up']:
+ monitor = m
break
else:
return '', []