Replies: 2 comments 1 reply
|
Thank you for reporting this issue. Root CauseThis regression is caused by a change in the default UART baud rate in the OpenThread POSIX platform configuration. In commit 16bafadb5, the default baud rate for the Spinel UART interface was changed from The refreshed snap (revision 85) was built with a version of OpenThread containing this change. Because your Radio URL configuration does not explicitly specify a baud rate, WorkaroundYou can resolve this issue on the refreshed snap by explicitly defining the baud rate in your Radio URL configuration. Update your configuration (typically in (Note: The snap is maintained by Canonical IoT Labs, and users employing hardware that runs at the legacy |

Uh oh!
There was an error while loading. Please reload this page.
AI summary:
otbr-agent fails to start after snap refresh on ConBee III RCP;
revert restores service
Environment
Snap: openthread-border-router
Channel: latest/beta
Installed revision before revert: 85 (thread-reference-20250612+snap)
Refresh happened about 4 days ago
Host: Ubuntu system running snapd
Hardware: ConBee III used as OpenThread RCP
Another ConBee III on the same host is used for Zigbee/Phoscon and works fine
Summary
After a snap refresh, otbr-agent repeatedly fails to start on a ConBee III RCP with:
P-SpinelDrive-: Wait for response timeout
Platform------: Init() at spinel_driver.cpp:87: Failure
The service exits with status 1 and systemd keeps restarting it. Reverting the snap immediately restores service.
Symptoms
After the refresh, the service starts, opens the configured serial device, and then times out waiting for a response from the radio co-processor.
Relevant log excerpts:
otbr-agent[...] [NOTE]-AGENT---: Radio URL: spinel+hdlc+uart:///dev/serial/by-id/usb-dresden_elektronik_ConBee_III_DE03311466-if00-port0
otbr-agent[...] [W] P-SpinelDrive-: Wait for response timeout
otbr-agent[...] [C] Platform------: Init() at spinel_driver.cpp:87: Failure
systemd[1]: snap.openthread-border-router.otbr-agent.service: Main process exited, code=exited, status=1/FAILURE
The service then enters a restart loop.
Diagnostics performed
The device is present and the serial path exists.
ls -l /dev/serial/by-id/
Output includes:
usb-dresden_elektronik_ConBee_III_DE03311465-if00-port0 -> ../../ttyUSB1
usb-dresden_elektronik_ConBee_III_DE03311466-if00-port0 -> ../../ttyUSB0
fuser shows otbr-agent is opening the device.
sudo fuser -v /dev/serial/by-id/usb-dresden_elektronik_ConBee_III_DE03311466-if00-port0
Output:
/dev/ttyUSB0: root 3740763 F.... otbr-agent
No USB disconnect/reset errors appear in dmesg for the device.
sudo dmesg | grep -iE 'ttyUSB0|conbee|usb|reset|disconnect|error'
The output only showed unrelated ACPI BIOS warnings and no evidence of USB transport failure.
Workaround / recovery
Reverting the snap fixes the issue immediately.
sudo snap revert openthread-border-router
After revert, otbr-agent starts normally again and the Thread RCP becomes usable.
Impact
This appears to be a regression in the refreshed snap revision. The hardware and USB path are fine, and the issue is reproducible after refresh and resolved by revert.
Additional note
The other ConBee III on the same host is used by Phoscon/Zigbee and continues to work normally, so this does not appear to be a general USB or host issue.
Request
Please investigate whether the refreshed openthread-border-router snap revision introduced a regression in RCP startup or Spinel/UART communication with ConBee III devices.
All reactions