Closed djansen1987 closed 2 months ago
agners
commented
7 months ago There are several similar reports with USB devices going missing in this issue tracker, e.g. #2995.
Did this start with recently? E.g. can you try to downgrade to 11.1 using ha os update --version 11.1?
djansen1987
commented
7 months ago Hi @agners, thanks for you quick reply. Unfortunally when i created my post 2 months back i was running version 11.1 and it might even already been in an OS version before that. Strange part is it could be running days, sometimes weeks without an issue. But sometimes also twice a day. Nothing noteworthy happening that can relate the events for me. I will maintain looking in the logs when it occurse again, but this makes the wife very unhappy...
agners
commented
7 months ago @djansen1987 you can also go further back then 11.1, e.g. 10.5. You are certain that the same hardware combination used to work reliably at one point?
From the host logs you've provided (which include the kernel logs) it seems they are cut-off or really doesn't have any issue reported (no log entries from January 16th).
It does sound like it is OS related indeed since the device doesn't come back even after unplugging/replugging, but I'd expect kernel errors in this case.
Can you double check the kernel logs using ha host logs --identifier kernel -n 10000? Especially for that timeframe the Core and Z2M reported issues?
djansen1987
commented
7 months ago I am sure the config has worked correct at some point. Let me know after viewing the logs if you still want me to revert to the old os version (what might i experience when reverting ?)
It seems like the host log time is 1 hour off (earlier) with the zigbee2mqtt log. So start looking in the host.log at:
Jan 15 08:51:12 homeassistant kernel: Bluetooth: hci0: command 0x200c tx timeout
Somapolice
commented
7 months ago I am having the same issue as you with a Dell Wyse 5070 running 11.4
First the bluetooth times out then it cascades to my USB ports timing out (I have a zwave controller, a zigbee controller, and a coral tpu connected via usb).
khayamgondal
commented
7 months ago Same problem here for last 6 months. Hardware is Intel NUC 5i5RYK. I have tried several Linux kernel, Ubuntu and Home Assistant versions but the problem persists. I have a Zwave and zigbee USB sticks connected. I guess it is either a problem with the USB hub (Bluetooth is internally connected to the USB HUB) or a power supply issue. I am using a generic 65W PSU. Note: I am using HA docker not HAOS
djansen1987
commented
7 months ago Hi @agners, did you find anything in the logs ? It occurred multiple times in the meanwhile. If need i can collect more logs
agners
commented
6 months ago The logs show various USB related issues, e.g. this:
Jan 15 08:53:06 homeassistant kernel: usb 2-7: reset full-speed USB device number 5 using xhci_hcd
Jan 15 08:53:21 homeassistant kernel: usb 2-7: device descriptor read/64, error -110
Jan 15 08:53:37 homeassistant kernel: usb 2-7: device descriptor read/64, error -110
Jan 15 08:53:37 homeassistant kernel: usb 2-7: reset full-speed USB device number 5 using xhci_hcd
Jan 15 08:53:53 homeassistant kernel: usb 2-7: device descriptor read/64, error -110
Jan 15 08:54:09 homeassistant kernel: usb 2-7: device descriptor read/64, error -110
Jan 15 08:54:09 homeassistant kernel: usb 2-7: reset full-speed USB device number 5 using xhci_hcd
Jan 15 08:54:14 homeassistant kernel: xhci_hcd 0000:00:14.0: Timeout while waiting for setup device command
Jan 15 08:54:20 homeassistant kernel: xhci_hcd 0000:00:14.0: Timeout while waiting for setup device command
Jan 15 08:54:20 homeassistant kernel: usb 2-7: device not accepting address 5, error -62
Jan 15 08:54:20 homeassistant kernel: usb 2-7: reset full-speed USB device number 5 using xhci_hcd
Jan 15 08:54:31 homeassistant kernel: xhci_hcd 0000:00:14.0: Abort failed to stop command ring: -110
Jan 15 08:54:31 homeassistant kernel: xhci_hcd 0000:00:14.0: xHCI host controller not responding, assume dead
Jan 15 08:54:31 homeassistant kernel: xhci_hcd 0000:00:14.0: HC died; cleaning up
Jan 15 08:54:31 homeassistant kernel: xhci_hcd 0000:00:14.0: Timeout while waiting for setup device command
Jan 15 08:54:31 homeassistant kernel: usb 2-1: USB disconnect, device number 2
Jan 15 08:54:31 homeassistant kernel: usb 2-1: failed to send control message: -19
Jan 15 08:54:31 homeassistant kernel: ch341-uart ttyUSB1: ch341-uart converter now disconnected from ttyUSB1
Jan 15 08:54:31 homeassistant kernel: ch341 2-1:1.0: device disconnected
Jan 15 08:54:31 homeassistant kernel: usb 2-3: USB disconnect, device number 3
Jan 15 08:54:31 homeassistant kernel: ftdi_sio ttyUSB0: error from flowcontrol urb
Jan 15 08:54:31 homeassistant kernel: ftdi_sio ttyUSB0: FTDI USB Serial Device converter now disconnected from ttyUSB0
Jan 15 08:54:31 homeassistant kernel: ftdi_sio 2-3:1.0: device disconnected
Jan 15 08:54:31 homeassistant kernel: usb 2-4: USB disconnect, device number 4
Jan 15 08:54:31 homeassistant kernel: usb 2-7: USB disconnect, device number 0This is most likely a Linux kernel issue. Given that USB seems to work rather reliable for must users, it seems to be some interaction with newer stable kernel releases and your particular system.
We are soon going to update to Linux 6.6, initial development version on the dev channel will soon follow once https://github.com/home-assistant/operating-system/pull/3063 is merged. Maybe this new major release helps? :thinking: :crossed_fingers:
djansen1987
commented
6 months ago Okay, goed to hear it is still moving forward. Will wait for the new kernal update to be production ready
henriklund
commented
6 months ago @djansen1987 what Zigbee dongle are you using?
sairon
commented
6 months ago @djansen1987 OS version 12 with the updated kernel is in the stable channel, please let us know if it fixed your issues.
djansen1987
commented
6 months ago @henriklund i have a "zzh! CC2652R Multiprotocol RF Stick" with the latest firmware. I do have to say i has been a while for it to break. What i also noticed but it might be a coincidence. That when it happend the stick was in a continues "permit join" state as i forgot to turn it of a couple of time. Did not yet had the time to see if i could reproduce.
PS: Maybe unrelated but, last night my homeassistant web interface was not responding so was the app. I was able to run turn off/on devices via google voice assistant -> nabu casa. When i went to the console screen on my NUC i saw the following:
(ignore the ```````` on the command line as that was me pluggin in a keyboard)
henriklund
commented
6 months ago My observations are somewhat similar and then again... running with a ConBee II and Aeotec ZWave Stick for a few years. Until recently I (also) ran Home Assistant from a NUC. Although all devices worked like a charm after a reboot, I would frequently (within days or 1-2 weeks) experience the USB sticks getting disconnected. On a few occassions they reconnected, but the predominant experience was that they remained disconnected. This was a bit of a nuisance for which I had to create a few automations to monitor the state and implement responses (restart integration or even reboot the host). Two months ago I decided to run HA virtualized (on Proxmox) as means to better utilize the NUC capacity. All hardware remained the same. I noted an improved handling of the hardware (less frequent disconnects of USB), Only problem now was that the Zigbee network (through ConBee II) started becoming unstable (I saw errors in the logs that indicated some capacity constrain / device fault). I opted for replacing with ConBee III (less hassle on migration) and behold, no more USB disconnects. To say I was pleased is an understatement.
That is until I started rebooting the NUC. Now the ZWave chimes in with a connect, only to disconnect 20sec later:
Ejecting and reinserting the USB (from Proxmox by removing / adding the device) works without problems and afterwards everything is stable.
(this was also the case before the latest update).
sairon
commented
6 months ago @henriklund So do I understand it correctly that with bare-metal install, you had more disconnects, and with ConBee III in a VM, it does not happen anymore?
About the first log capture, if you don't filter for USB, is there anything around that USB disconnect event that would justify it?
henriklund
commented
6 months ago @sairon you are correct. My infrastructure seems far better off now (running virtualized and with ConBee III) than two months ago when it ran bare-metal. As soon as all is up and running everything appears stable; nothing in logs, no sudden lights failing to turn on / off. The only hiccup left if the odd disconnect of the ZWave USB during boot. And even here I yesterday saw the odd case (when trying it out) where everything started up and worked; it happened once during 5 tries.
This is what I see relates to discovering USBs:
Then a lot of logging related to apparmour, docker, network, CIFS before this:
henriklund
commented
5 months ago @sairon, I do not know if my error qualifies as a new issue (I do not want to hijack this one). If so, please let me know.
However, I had a strange observation today. I was trying to experiment with some simple monitoring of my ZWave. While doing so I restarted ZWave JS UI (to simulate loss of ZWave). What intrigued me was that the first restart went fine while the second caused the USB to become disconnected. Time between restarts did not seem to influence the outcome (whether it was 2 minutes or 5); the USB was no longer visible in /dev/serial/by-id/ and last entry in dmesg was a USB disconnect. A quick remove and add in Proxmox (ie. a virtual inject insert) and USB was back.
henriklund
commented
5 months ago Worth mentioning; this appears to a consistent behaviour. Second restart makes the USB disconnect.
sairon
commented
5 months ago @henriklund Problems with the ZWave sticks are tracked in #2995, and there is a fix/workaround in 12.1.rc1. Let's keep this issue only for problems not related to ZWave sticks, and continue in discussion in the other one if the ZWave problems persist for you after upgrading to 12.1.rc1.
github-actions[bot]
commented
2 months ago There hasn't been any activity on this issue recently. To keep our backlog manageable we have to clean old issues, as many of them have already been resolved with the latest updates. Please make sure to update to the latest Home Assistant OS version and check if that solves the issue. Let us know if that works for you by adding a comment 👍 This issue has now been marked as stale and will be closed if no further activity occurs. Thank you for your contributions.
acvigue
commented
1 month ago Seeing the same issue with a 5th gen i5 nuc
henriklund
commented
1 month ago @acvigue, this issue is closed. I think it would be better to raise a new request to track. I decided to do something in an attempt to ‘fix’ this since I ran HA inside a ProxMox on my NUC. So far it has not reoccurred since. The rough outline was that I:
Martinvdm
commented
3 days ago Although this is an old topic, it seems that i have the same combination. Conbee II, Aeontec Zwave. After enabling Bluetooth integration, it occasionally disconnects USBs.
Describe the issue you are experiencing
My zigbee2mqqt was failing once in a while. Rebooting the system resolved it. Today it happened again so i started digging through the log files, there i found in de debug log of zigbee2mqtt that the USB Dongle was suddenly disconnected. Looking into the HA core logfiles i could see at that time not only my zigbee dongle became unavailable, but also my Bluetooth chip and DSMR P1 USB dongle.
If i go into the terminal and do a less /dev/tty… there is no USB device available.
I Also have a Intel Nuc, NUC5i3RYH with HA: Home Assistant 2024.1.3 Supervisor 2023.12.0 Operating System 11.4 Frontend 20240104.0
The USB dongle is also extended with a sort usb extend cable (20 cm)
If i go the all hardware under system i see a couple of usb ports listed as:
This is wat the core logfiles tell right before failing all usb ports:
This allmost looks like drivers failing or something, who can help?
Post on the community with other users with the same issue: https://community.home-assistant.io/t/usb-ports-lost/446289/9?u=djansen1987
What operating system image do you use?
generic-x86-64 (Generic UEFI capable x86-64 systems)
What version of Home Assistant Operating System is installed?
11.4
Did you upgrade the Operating System.
Yes
Steps to reproduce the issue
Anything in the Supervisor logs that might be useful for us?
System information
System Information
Home Assistant Community Store
GitHub API | ok -- | -- GitHub Content | ok GitHub Web | ok GitHub API Calls Remaining | 4924 Installed Version | 1.33.0 Stage | running Available Repositories | 1377 Downloaded Repositories | 11Home Assistant Cloud
logged_in | true -- | -- subscription_expiration | August 7, 2024 at 2:00 AM relayer_connected | true relayer_region | eu-central-1 remote_enabled | true remote_connected | true alexa_enabled | false google_enabled | true remote_server | eu-central-1-3.ui.nabu.casa certificate_status | ready instance_id | a00d7c3253584417bca037041fbc4817 can_reach_cert_server | ok can_reach_cloud_auth | ok can_reach_cloud | okHome Assistant Supervisor
host_os | Home Assistant OS 11.4 -- | -- update_channel | stable supervisor_version | supervisor-2023.12.0 agent_version | 1.6.0 docker_version | 24.0.7 disk_total | 458.4 GB disk_used | 88.9 GB healthy | true supported | true board | generic-x86-64 supervisor_api | ok version_api | ok installed_addons | Samba share (12.2.0), Home Assistant Google Drive Backup (0.112.1), Node-RED (17.0.2), Mosquitto broker (6.4.0), motionEye (0.20.0), Terminal & SSH (9.8.1), ESPHome (2023.12.5), AirCast (4.1.1), emby (4.7.11), Zigbee2MQTT (1.35.1-1), Advanced SSH & Web Terminal (17.0.3), Matter Server (5.0.2), AirSonos (4.1.1), FTP (5.0.1), Nextcloud (28.0.1-6), Music Assistant BETA (2.0.0b82)Dashboards
dashboards | 2 -- | -- resources | 6 views | 31 mode | storageRecorder
oldest_recorder_run | January 5, 2024 at 5:38 PM -- | -- current_recorder_run | January 15, 2024 at 10:22 AM estimated_db_size | 2373.16 MiB database_engine | sqlite database_version | 3.41.2Spotify
api_endpoint_reachable | ok -- | --Additional information
No response