-
-
Notifications
You must be signed in to change notification settings - Fork 30.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Client unable to keep up with pending messages. Stayed over 512 for 5 seconds. The system's load is too high or an integration is misbehaving #82562
Comments
I have the same issue and in addition to it the following error: Logger: homeassistant.components.websocket_api.http.connection At the same time Node-Red is having issues of repeated disconnection/reconnection, seemingly this has something to do with it. When it first appeared I was trying the new Bluetooth related capabilities, and iBeacon tracker created a lot of related entries. HA version is 2022.11.2 |
same issue here , also Node-Red is disconneting and connecting all the time... HA/Node-Red very very unreliable for the last weeks... |
Same issue here Logger: homeassistant.components.websocket_api.http.connection [281472552731696] Client unable to keep up with pending messages. Stayed over 512 for 5 seconds. The system's load is too high or an integration is misbehaving |
I also have this issue on 2022.1.4 but I do not have Node-Red like some of the others
|
As others are saying in other forums, deleting entities not used any more, cleaning up in Node-red, etc seems to help for a while, then it comes back, and I am really interested how to catch the guilty process or integration... |
One of my home assistant installations is having the same issue.
|
Any progress on this one? I am facing the same issue. It appears to be impacting the MQTT messages as well (i.e. devices not being able to connect) . I do have close to a hundred devices on localTuya and it is core to the automations. The system is running on a VM with 8 cores and 10 GBs memory - so grunt is not an issue. |
+1 for about 3 months now. It was making my RasPi4 unresponsive at 26% processorUsage and requiring 3 reboots to clear the problem. I have now upgraded to a OptiPlex 7050 with i5 processor and now, after less than 12 hours since reboot, 12 million error messages logged. "Not happy Jan". Thankfully the processorUsage shows 12% and doesn't bog the system down excessively like on the RasPi.
sensor.connected_clients is showing 5 |
Is there any way of diagnosing what are making the problematic calls? |
Same issue since yesterday. HA restart doesn't help.
The error from subject is always coming together with:
I'm not sure, if it has something to do with the subject, but at the same day I got info message for one Hue bulb: An update via Z2M integration doesn't work: |
+1. A restart fixes it for me. |
Genuine question: Does the restart fix it or simply clear the log (and it reoccurs later)? |
It fixes it temporarily (hence the +1 reply above). Sometimes it will be days before it reappears. I only notice it since my syslog server sends me an alert of the repeating messages. Still looking for a fix. |
cheers. |
This may not be the case for all users. In my case I now think the problem only occurs directly after a reboot, subsequently stopping - suggesting it was some artefact of the system catching up at startup. |
It would be very helpful to get some sort of pointer to WHICH part is causing the problem. I get this alert to intermittently. And yes a reboot mostly sorts it out for a while. But still would like to keep my log 'clean' of errors as much as possible. This one is partly unfixable atm |
I got to this thread when I encountered the same error and what was interesting was that I have been encountering similar symptoms with my HA for a long time, where it disconnects and reconnects constantly, and most annoyingly refreshing my Node-Red and destroying all the work that I did since I last deployed it. Along the way - I wont go into all the integrations that I tried to take out or disable to find my own culprit. To be honest, I thought it was associated with the TUYA integration, but nothing changed with disabling or removing that integration. When I read through @ss89 comment to see what integrations were common, of all things, I tried to disable the SUN integration under pure desperation. I could not believe my eyes. All my disconnects stopped! While I am not sure this will be the fix for other, I wanted to contribute to this and see if it helped anyone else. I really wish there was a better way to diagnose what integration was causing my issues, but I was just floored that this could have caused my issue. Yes - I did move to the latest HA version 2023.1.0, but once I did, I got other errors and I solved those, but still had the disconnect issue. I kept hoping that each new release would solve the issue, but again no. So I decided to spend some time to diagnose it (AGAIN!) and then I looked over @ss89 list of integrations and decided to just disable my SUN integration. Now my new issue is that I use the SUN integration for monitoring the sun angles and elevation to adjust the shades in my house, so I will be heading down the path of, removing it all together and then seeing what happens when I install it again. Again - I am offering this as something to try. I will post as I continue to find anything new - good or bad. |
Thanks for your Hint @MihleTX ! |
I haven't seen the issue since. |
New with HA and on my way to migrate my stuff over. |
Can you still use sunrise/sunset conditions in automations if the sun integration is disabled? |
Guess not, but I'm only using my light sensor for deciding when to turn lights on/off. |
I use Node Red and BigTimer - and it is not leveraging the SUN integration for it. |
I used to have all my window shades integrated using the SUN integration and the sun position to determine when to open and close, but I switched over to light sensors in the rooms and that worked better anyway. |
Tried to disable sun integration but had the same error again this morning at 7:57 |
the same here. Newest HA version L``` [548012469472] Client unable to keep up with pending messages. Stayed over 512 for 5 seconds. The system's load is too high or an integration is misbehaving
|
scratch that - I still get them. |
Some cases can be caused by a large number of entities. The linked pr should improve the situation when you have 2000+ entities. If you still get the message after 2023.3.0 (2023.2.0 will not have the change) be sure to include a callgrind.out file from the profiler integration profiler.start service in any new issues |
OK...sounds good it will be solved! |
You can generate a callgrind with https://www.home-assistant.io/integrations/profiler/#service-profilerstart |
OK - so the file is not a solution for the problem, it's some kind of log/debug file. |
Disabling the Sun integration does not prevent this error occurring on my setup, under the latest version (no issues on 2022.11). I'm using HA Container on Synology NAS (DS920+ with 20GB RAM). Hugely frustrating, two and half months into trying to work through this, but glad I'm not alone! |
I upgraded to an i5 PC to make sure I had enough CPU for my 20 devices, and the number of errors only increased - so I believe there is something more serious going on. So I think it's a little premature to close the PR a month before we can test it. If only there was some clue as to what the offending messages are, where they are coming from, or where they are going to. |
I had this a couple of weeks ago. I can't pinpoint exactly what fixed it, but I suspect it was zigbee2mqtt. Make sure you have the latest version installed. |
I'm on 1.30.1 which I think is the latest. Unfortunately, I've still got the issue. |
I'm also on 1.30.1 (latest) and also still have the issue. |
It seems the problem still exists in 2023.3, but now the system slows to a crawl (even on my i5 PC) for much longer before reporting the error.
The good news is that the error message now gives information about where the issue lies, ... I think ... |
What's the full log entry (pointing to the where the issue lies)? |
There is the following comment in the "hints" section for this card, but not sure if it applies to you:
There is also a long-running bug report for "Client exceeded max pending messages". Again, not sure if it applies to you, but maybe worth look? It refers to problems when the card is permanently displayed (maybe on a wall-mounted device) or similar. |
There are many many bug reports for these error messages, which I have been looking at periodically since September ... and lots of guesses as to which is the "misbehaving" integration.
That was one of the things I looked into, and yes it does now appear to be my problem. I don't want the table constantly updating unless the user is actually looking at that Lovelace tab, so always_update is set false. And do_not_parse prevents the templates to insert the values of the sensors and attributes, no that defeats the purpose. At least the new error message does show that for me, the HTML tables are the issue, and I need to think or find another approach. |
The problem
In the logfile for the websocket_api I see the error below, I am unable to check what the cause of this error is:
Logger: homeassistant.components.websocket_api.http.connection
Source: components/websocket_api/http.py:157
Integration: Home Assistant WebSocket API (documentation, issues)
First occurred: 07:21:40 (1 occurrences)
Last logged: 07:21:40
[281472675383568] Client unable to keep up with pending messages. Stayed over 512 for 5 seconds. The system's load is too high or an integration is misbehaving
What version of Home Assistant Core has the issue?
2022.11.04
What was the last working version of Home Assistant Core?
No response
What type of installation are you running?
Home Assistant OS
Integration causing the issue
websocket api
Link to integration documentation on our website
https://www.home-assistant.io/integrations/websocket_api/
Diagnostics information
No response
Example YAML snippet
No response
Anything in the logs that might be useful for us?
Additional information
No response
The text was updated successfully, but these errors were encountered: