Integration fails to start up on HA restart (Public API) #90
Comments
|
Just confirmed - after around 30mins my energy monitor stops working on the internal API. I switched over to Public API around 3 hours later (without restarting) and it switched over and started working fine. The energy monitor works fine on the Public API. |
|
Thanks for the full logs. Line 216 has "API response 500". A HTML error code of 500 is the not-so-useful "Internal Server Error". Looks like an issue at the LW end. Can you post additional logs if it recurs? Maybe there's a pattern in which requests cause the error. |
|
I am intermittently getting an LW Plugin config error on HA restart. I restart HA every second day, and I see the error at least once a week. Manual re-adding the LW integration (without removing the old one) always solves the issue. I hope the logs help. |
|
@mydliar When the integration sends messages to the server, it waits up to 5 seconds for a reply and then times out. It looks like your responses are taking slightly around 5 seconds (e.g. message 1, sent 0939:33, received 0939:38) - normally this should be pretty instantaneous. Is your network under high load? In any case, this is a flaw in my code, it should recover from a timeout rather than crashing. I will fix imminently. |
|
@bigbadblunt - thanks for this. I am on a very fast 1Gb fibre broadband (and my home LAN is also running on 1Gb), but the NAS where I run the HA is under permanent network load from the CCTV camera feeds. That could explain it, though I never experienced latency or speed issues on the NAS. Thanks again for looking into it. |
|
So I've had my lightwave drop out many times, just kept forgetting to grab the logs before restarting. As far as I can tell I lose data updates to my entities at 21:11:52. At the same time, the debug logs (which have been very verbose up until that point) simply stop. I've attached a snippet from the timeframe: You can see the latest message happens at 21:11:52.575 for lightwave, then the next messages I have in the logs are from other integrations throwing some warnings. I don't see that this particularly helps, though! The only other thing I could see in the logs (although its from an hour earlier) was this: Which looks like it recovered OK to me. Hope that helps |
|
What do you mean by "drop out"? You say you lose data updates; can you still control the entities from HA? If so, this is a duplicate of #69 |
|
Ah, sorry, my bad. I commented on the wrong thing :( I've actually not been experiencing this restart issue for quite some time now. The only time I've noticed it has problems is when HA has started up before the LW hub is connected to the internet. In this case HA does not connect to LW (and nothing can be controlled) until I restart HA to reestablish the connection properly. I suspect that is a different problem to what I originally raised here. Personally I'm happy for this one to be closed, as I'm not having this problem any more. |
|
Looks like the original problem here is now fixed in #108. I am keeping open until I fix the recovery from timeout issue uncovered a few comments higher. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Recently when I've been restarting home assistant, I've been left with failures for the LW integration to start up.
HACS LW Plugin Version:
v3.5.15HACS Version:
1.26.2Home Assistant Version:
2022.7.5The way I've been able to get things going again is by turning off the Public API, restarting to get things going. Then waiting a little while until I hit a problem where my energy monitor no longer reports updates to HA (I should raise a separate issue about that I guess!) then switch to the Public API. I think that has been the process anyway, I've been doing a lot of option/restart twiddling recently, so will be paying more attention now to give a more accurate bug report.
Full logs attached
fulllogs.txt
The text was updated successfully, but these errors were encountered: