macromark said:
We use the standard Sigma library in our own controllers. However, our systems do have the capability to set the command routes (the precise path from the controller out to the device). In many cases, this feature alone will fix communication problems*. However, in this case, the network density is simply too low. if you can't communicate adequately to the outlet in the garage, there's virtually no hope of controlling the lock next to it. Lock communications are very picky and require clean command and return routes to work properly. I would recommend adding a few more receptacles or plug-in modules to boost network density. Then, I would recommend running a "full optimize" on the network to store return routes in the lock's built in module memory.
*Requires SmartStick+, Z-NET, Z-Wave.me UZB or Aeon Z-Stick GEN5.
The Leviton VRC0P (gen3) has that capability of manually setting a route, too. However, the manual setting, by itself, does not guarantee that the assigned route will be taken by the packet, as I understand. So, functionally, it's no different from each node having a collection of neighbors discovered by the controller during the "healing" process and stored in the device memory.
Besides, in my observations, a device does not really try an alternative route if there is no ack from the controller but blindly re-transmits two times and then gives up. It is always the controller that decides to try a new route if there is no ack from the device.
E.g.:
S - source id
D- dest id
(L) - length
-79/-80/-42 RSSI
3105010943b5 - thermo specific payload
A thermostat to the controller S D
2016-05-19T01:36:03.181737 homeid 4 13 (15) 3105010943b5 -79 65F temperature
No ack S D
2016-05-19T01:36:03.248765 homeid 4 13 (15) 3105010943b5 -79 Try again
No ack
Give up.
A new attempt after the temperature changes:
2016-05-19T04:23:11.120455 homeid 4 13 (15) 3105010942b5 -80 66F temperature
2016-05-19T04:23:11.125532 homeid 13 4 (10) 8c -42 Got an ack
Success
Perhaps, things improved in gen5, perhaps this sort of behavior(not trying an alternative route) is specific for my thermostats. Who knows ...