Hello,
i was wondering if the watchdog device can be exposed to a container by just passing the device /dev/watchdog to a container and additionally how is it possible to know inside a container the reason for the reboot if triggered by watchdog or otherwise.
Greetings @nkj,
I want to better understand your situation. Therefore, could you please describe what your use-case/requirements are here?
Do you want to use/interact the watchdog from within a container? Or is it just for monitoring purposes?
I can provide a bit information about how the watchdog is configured on TorizonCore. By default the watchdog is configured to be used by and monitor systemd. This does mean that the watchdog can’t be “grabbed” by other applications since it’s being used already. However this behavior can be modified depending on what you need.
Let me know what your use-case case is and we can discuss this further.
Best Regards,
Jeremias
Ok if the base torizon system uses the watchdog than its good. the next question was us would be is there a way to know the reason of the restart of the computer for e.g. ota update, watchdog or normal user restart ?
Hi @nkj !
This could help you to find out the reboot reason: Persistent Journald Logging | Toradex Developer Center
Best regards,
Hello,
I have a question related to this post. Is there a way to use sd_notify(“WATCHDOG=1”) inside a container or other method to reboot in case my program hangs?
Best Regards,
Francisco
Hi @franc !
Thanks for joining our community. Can you please open a new thread about it as it seems to be a little different from the original question and other users may take profit from this discussion?
Best regards,
Guilherme
