Plugwise_usb-beta v0.40.4 available for testing

[bouwew]

Please test the latest backend update, v0.31.1b0: you can switch to using this test-version by updating the plugwise-usb version in the plugwise_usb-beta custom_component by changing the version-number in the manifest-json file from 0.31.0 to 0.31.1b0.
Save the updated file and then restart HA Core.

Please post your results here.

[bouwew]

After running for 2 hours I see these warnings:

I did see these warnings with the previous version of the plugwise-usb backend as well. So everything still "normal".

[dirixmjm]

Yes, the fix is not related to these messages.
However, the occurrence of these messages should be lower because the stick is not flooded with CircleEnergyCountersRequests anylonger.

[dirixmjm]

@herwinjan you can check inside python-plugwise-usb/plugwise_usb/node/circle.py on line 618.
If you have the latest update it should say "> datetime.now()" instead of "< datetime.now()".

[herwinjan]

check, thanx. My network is not OK I think. not all devices are online. Non of them gives Energy consumption today, some ging a switch (but I cannot switch) and some of them gives the watt usage.

[dirixmjm]

It can happen that after a restart it takes a while for the circles to come online.
Can you do a fresh restart, and then generate a debug log running some 10 minutes? I'll try if I can find anything.
I can hardly imagine that this fix is related, however to be sure you can also try and revert back to the 0.31.0 version and check.

[dirixmjm]

One more comment, please try a power cycle of the Stick before you restart. If the Stick is flooded it requires a power cycle to reset the memory.

[herwinjan]

Here is the log after the restart and power off of the stick.
home-assistant_plugwise_usb_2023-12-29T13-50-48.744Z.log

[dirixmjm]

You tried a power cycle of the Stick before restart?

[herwinjan]

Hi. yes! 1 minute off the USB port.

[dirixmjm]

I don't see a CircleSwitchRelayRequest in the log. Most likely this plug was not marked available?
According to the log only 1B8Bc and B996D are available.
All other circles are not responding to the initialisation ping, most likely you need to wait longer for them to start responding, maybe the network needs to update routing.
Can you let the network sort itself overnight and come back?

[dirixmjm]

@bouwew @herwinjan I found that the logic in the if statement line 611-613 still has a flaw resulting in flooding. I'll need to sent another PR shortly after I have done some more testing.

[dirixmjm]

Just to prevent interference, can you update the D3317E circle?

I would not know what Source does differently, if I compare the comport log to what the module is doing it is very similar, except for the handling of the energycounters maybe, Source does not really care if intermediate logvalues are missing.
If Source is really performing better in your case, you could consider to log the comport communication with your network.

[dirixmjm]

I made the communication a little quieter still. But not sure it will help much.
Can you try (three files have changed, please put them at the same spots as in the zip-file inside the plugwise_usb folder.

plugwise_usb_1201_1228.zip

[dirixmjm]

@herwinjan I think I have found the main problem we are facing, there is a bug in the sequency-ids.
That does not mean that I have a solution yet (and next week I'm on a business trip). So for now probably not much progress.

[dirixmjm]

If you want you can give this controller.py a try, it is a hack but it seems to work on my side.

controller.zip

[dirixmjm]

After running for a night, the network is more stable, power_consumption_{hour,today} are working fine.
however energy_consumption_today still gets stuck.

[dirixmjm]

[CoMPaTech]

It looks like your are replying to something, or you are answering in the quote? (context seems lost)