#here s the log file starting from a
novel kindle
You'll want to tone down the reporting on some of your power meters:
https://zwave-js.github.io/node-zwave-js/#/troubleshooting/network-health?id=optimizing-the-reporting-configuration
right when you attempt to lock the first lock, node 7 bursts 4 measurements over half a second. The first one collides with the lock command, zwave-js backs off 100 ms, then the next one comes and collides again, zwave-js backs off for ~1.
Then it requests a nonce from the next lock, which takes 3 seconds to reach (not sure why, but a low link quality might be a reason).
That command goes through, but the following communication attempt with the third lock is again slowed down by a power meter reporting, this time node 10.
On top of that, the locks don't seem to have the best connection strength. For 29, RSSI values between -90 and -100 are reported repeatedly - that is close to the background noise for many.
Make sure you do this:
https://zwave-js.github.io/node-zwave-js/#/troubleshooting/connectivity-issues?id=general-troubleshooting
and then do a health check (https://zwave-js.github.io/node-zwave-js/#/troubleshooting/network-health?id=testing-the-connection-strength) on your nodes.
safe zephyr
Thanks so much Al-mighty!
The power reporting devices I can understand flooding the network
But why don't the locks use the repeaters all the time to improve the signal quality?
The repeaters themselves can be high 60 low 70, much better than 90 to 100
novel kindle
But why don't the locks use the repeaters all the time to improve the signal qua...
That is the mystery of Z-Wave routing. If your mesh is strong enough, this shouldn't matter
I'll add a way to use fixed routes at some point but ensuring a strong mesh is always good
safe zephyr
All my power metering devices are Aeotec. To lesson the network load, I just moved 2 parameters from node 7 (my whole house clamp energy meter) from 10 seconds to 120 seconds, leaving just A and W reporting every 10 seconds now.
Node 10 might have been a fluke, that is my smart switch 6, reporting for that is 90, 180 and 1 day for A, W, V and kWh
novel kindle
90 what?
I recommend not using time-based reports if possible in any way. If you have the options (Aeotec is quite configurable), prefer reports on significant changes (like +/- 0.5 A or +/- 5W)
safe zephyr
seconds
ok, that is a good point. I can look to report changes instead of time-based reports
safe zephyr
Hey @novel kindle, with regards to my comment in the beta forum above, before I mess things up, should I just delete the firmware entities that are not now available?
What is interesting is that I if go to z-wave js and click on the device's OTA updates...I have the ability to check for firmware updates, and when I do, it say none-available. but the entity in HA is still showing unavailable.
looks like it broke on B6
its not all the devices, just some
safe zephyr
no problem maybe I can ring @calm zodiac in when he has a chance
calm zodiac
no problem maybe I can ring <@490222347895439383> in when he has a chance
if your device does not have the firmware update metadata CC, we no longer create the update entity
that was a change in b6
those devices can't be updated
safe zephyr
@calm zodiac great, thanks for clearing that up! should I just delete the entity now?
it just appears as "unavailable" now
and once deleted, that should clear the log issue (I hope) because it approaching 100 occurrences (while on b6). I'm on 2022.9 now.
calm zodiac
yes you can delete it