Open rwkeane opened 3 months ago
I guess it's probably because otbr-web
didn't wait for the otbr-agent
to be ready, which it should have done.
Running systemctl stop otbr-web and then systemctl stop otbr-web to restart the service fixes the issue, and it runs without any errors.
I'm curious why you needed to stop it twice (or is it a typo? π ).
It's possible they are all fighting over the same mDNS name, and the probe -> retry -> repeat... name resolution process is taking a long time
Do you have logs supporting this? As far as I know this should rarely happen because the mDNS service instance name of a device ends with last few digits of its extended MAC address.
I'm curious why you needed to stop it twice (or is it a typo? π )
That's just a typo :)
Do you have logs supporting this?
It was only my guess based on Result of publish meshcop service OpenThread BorderRouter #55D9._meshcop._udp.local: OK
not occurring until after the otbr-web
had already launched (and failed)
The error [ERR ]-WEB-----: OpenThread daemon is not running
should be transient and will be automatically resolved when otbr-agent
is launched. I just tried launching otbr-agent
after otbr-web
and it's working fine.
Describe the bug A clear and concise description of what the bug is.
I have OTBR currently running on an RPI device (both 3 and 4 repro this issue). I am able to successfully start otbr-agent and it runs fine (including when started automatically as part of the boot process), but otbr-web fails after boot with error
[ERR ]-WEB-----: OpenThread daemon is not running.
Running
systemctl stop otbr-web
and thensystemctl stop otbr-web
to restart the service fixes the issue, and it runs without any errors.To Reproduce Information to reproduce the behavior, including:
Expected behavior A clear and concise description of what you expected to happen.
The web service should correctly delay starting until the agent is running.
Console/log output If applicable, add console/log output to help explain your problem.
on boot:
otbr-web (via
systemctl status otbr-web.service
):otbr-agent (via
systemctl status otbr-agent.service
):after restarting otbr-web with commands earlier (ignore the timestamps):
Additional context Add any other context about the problem here.
I've got 3 OTBR devices (RPI3, RPI4, mini computer w/ home assistant) as well as an official Thread Border Router (Nest device) running on my wifi network (though no thread network as of yet). It's possible they are all fighting over the same mDNS name, and the probe -> retry -> repeat... name resolution process is taking a long time