Not working since v0.9.9.18 #280
Comments
|
I have the same problem. The entity sensor.today_s_pv_generation and sensor.today_battery_charge are not updated always after midnight. |
|
@Legnava89 this is probably the problem of #279 #281 @kolossboss Could you please set the log level to DEBUG and provide the logs ? |
|
@kolossboss Could you please enable debug log and attach the log output ? |
|
Hi, thanks for reply. Sorry for late response. |
|
Hi, today I enabled the debug log and installed the latest version. thats the output. This are the diagnostic informations |
|
Hmm, that is odd. It seems you are receiving truncated/incomplete responses. |
|
Yes, |
|
There were changes in the code, but the log files look like problem is on network layer, not application. Did you try also the v0.9.9.19 version? |
|
I always have problems with the network connection. All Goodwe entities, shows sometimes unavailable for a short time. In the past I played a lot with these retry and timeout settings and scan rate 5 and timeout 30 works ok^^ |
|
Ok I tried again v0.9.9.19 with the new scan rate and timeout settings. |
|
That's unfortunate, since I do not have an explanation for that. |
|
The hardware, HA is running on? Yes, I think I can run a Python script. |
|
Ok, let's try to isolate the problem. Edit: Actually, no need for git - just download the sources directly - https://github.com/marcelblijleven/goodwe/archive/refs/heads/master.zip |
|
Hi, Thats the output: |
|
I've just release both library and HACS integration with new support for fragmented packets. |
|
Hi, Thats the log. |
|
I installed v0.9.9.21.
Log: 
|
|
Aaaaaaagh. |
|
Sorry man...its still not working 🙃 I can now configure the Integration. I disabled TCP keep alive. I also changed Inverter family to "EM". But the family is not saving. It always goes back to "ES". 
|
|
There's new version out there, please re-test. |
|
Sorry still not working with 0.9.9.23. Its not possible the change Inverter Family to Thats the log: 
|
|
ES and EM is the same stuff, so not issue. |
|
From version 0.9.9.18 I have problem with 10ET, that after 15-30 minutes integration terminates with this error. Every 30 minutes I have to reload config or restart HA. Error doing job: Exception in callback _SelectorDatagramTransport._read_ready() During handling of the above exception, another exception occurred: Traceback (most recent call last): |
If there is any way I can help you, feel free to ask. I really need this integration.😬 |
|
i have tree gw5048es . i have the same poblem i have downgrade to 9.9.22 its running now for more than 12h. |
|
@jiriknez Your issue seems to be of different kind. |
|
@kolossboss I cannot find anything in the code which would explain this behavior. |
|
I'm not exactly sure what you mean. I'm not so familiar with github. Also, I have no clue what to do... |
|
I will give you exact instructions, don't worry. |
|
Yeah, but the code which is in HACS will one day get into core HA, so will have the problem then until it gets fixed ;-) |
|
@mletenay Maybe at head I also added the whole file. Maybe you can finde something. Edit: It start to get better at 21:27:37: Maybe that helps :-) |
|
Hmm, I'm afraid I do not see any relevant correlation there. You receive broken/truncated UDP packets and I have no idea why. |
|
Does it makes any sense to go down any further with the commit? Its hard to believe that's an only network issue, because v0.9.9.17 is working just fine :-) Today I updated to v0.9.9.26. It takes a long time until the integration is initialized, but then I get partial values. But the update rate is 3-5 minutes, and sometimes the entities are unavailable for several minutes. Is there anything else I can do to help you?
|
|
I see but I do not understand why. |
|
Sorry for late response. Yes, I tried all settings. |
|
Yes, it's named that way and it is meant mostly for TCP. |
|
I tried with Keep Alive on. |
|
Just for the record, KeepAlive ON is new behavior, with KeepAlive OFF it should behave (for UDP) the same way as in 0.9.9.17. |
|
yes should😁 For me: 0.9.9.26 not working with keep alive off
|
|
@kolossboss |
|
Hi, Since the beginning, I have had connection reliability problems with my inverter (losing WiFi connection indicated by a blinking LED, SolarGo sometimes not working, etc.). So, I contacted Goodwe Support, and they installed the LAN module instead. It has gotten much better, but it’s still not perfect. I thought these reliability problems were by design and not a real issue, because the data is sent to the SEMS portal just fine. Maybe there is a general problem with my inverter... but why does v0.9.9.17 work well enough? In the web interface, the firmware shows V1.0.14. |
|
The WiFi module is running very stably for me. Despite being polled every second by Home Assistant, SEMS is also being filled with data. Sometimes the WiFi LED flashes and reports that it has no connection to SEMS. However, this only happens very rarely and resolves itself. I cannot see any failures in SEMS, and Home Assistant is still receiving current data without any problems. The inverter firmware does not seem to be up to date in my opinion. I have a 1314J (which is also on Goodwe's GoogleDrive). 1.0.14 is only the version of the web interface and not the module firmware. I don't currently know the version for the LAN stick. For the WiFi module it is 1.1.3.10. |
|
Hi, Edit: |
|
There are three updates for the inverter: ARM, DSP, communication module (WLAN 1.0, WLAN 2.0 or LAN) |
|
The Bluetooth stick is included with some series. Unfortunately not with the EM series. |
|
I do not have an usb dongle. I'm from germany (plz 212xx 30km südlich von Hamburg) |
|
Yes, we should find a solution. |
|
Yes, Thanks a lot. |
|
give ist an disscord server platform for the integration ? |
|
Just to add that version 0.9.9.27 works fine for me and I am not seeing the issues mentioned above. I recently switched from the official HA version to experimental as I was not able to set Eco mode. Everything works really well with 0.9.9.27 for me. As I am also having an EM inverter justed wanted to share my setup: GW5048-EM Would strongly suggest to apply for a firmware update with GoodWe support. They can be a little reluctant I learned myself, just keep asking :-) . |
|
... just to let you know, I'm still trying to figure out what the heck is going on there. |
|
Hi, thanks for all the help. I also get the Bluetooth USB Dongle from a very kind guy here Germany (thanks @smileyman-1 ). I tried every to update to the latest firmware. But in SolarGo, the update menu was not displayed, and in PVMaster, I couldn't select the firmware or it wasn't recognized when I placed it in the share folder. I tried it with both iOS and Android. I contacted GoodWe support, and they opened a ticket to upgrade the inverter's firmware. However, I have not received a response so far. @smileyman-1 also sent me a WiFi stick with firmware version 1.1.2.1. Today, I installed this stick, and it's working great so far! No issues with v0.9.9.27, and I'm getting new sensor updates every 10 seconds (or even every second) in HA. I've never had such a stable connection before. So, my LAN stick is definitely also an issue. Now I'm waiting for the inverter update, or maybe I'll find a way to upgrade the LAN stick with a better working firmware. edit: |
|
Just for the record: my "WiFi / Lan Kit 1.0" is on firmware V1.2.1.19 (Boot:V1.1.4) . No idea if this is the latest version. Anyway - it runs just fine with my setup. |
|
So, I have tested it a lot today and came to these conclusions:
|
|
I can confirm most of the points. |
But that is OK in my opinion - update frequency of SEMS depends on when the inverter decides to send data and is configured by Goodwe. It has nothing to do with HA polling. |
|
@janchlebek SEMS update frequency is controlled by register 47006 (default 12, *5s = 1m) and no, for some users out there, when HA polls too frequently, SEMS somehow stops receiving updates. |
Ah, did not know this is configurable. Thanks for clarifying. |
Generally, these modules (and the ne WLAN-Kit 2.0) exist: GA10081-22-00P is connected to my Goodwe GW3648-EM. - SEMS works perfectly even with a one-second update interval. GA10081-44-00P is connected to my Goodwe GW10KT-DT. - SEMS works perfectly from 180 seconds. Between 131 - 180 seconds, only some data is transferred to SEMS (Inverter is partially offline. The last values, etc. are missing in the chart.). Between 1 - 130 seconds, no data is transferred to SEMS. |
Hi,
for me, the integration is not working since [v0.9.9.18].
I got either no data and the entities shows
unavailableand the Integration is initializing or the entities shows a value but did not update for hours.If I went back to v0.9.9.17 everything went normal again.
I'm running an GW3648-EM Inverter.
Not sure, how I help solving this issue.
Thats the only Log entry:
Error fetching GoodWe data: ('No valid response received even after 6 retries', 4)The text was updated successfully, but these errors were encountered: