Turris Mox powers down

I’ve got a Mox A setup as a wireless access according to the instructions. Seems to work fine, red light flashes, I can use the access point to connect to the LAN and the internet via the main router (a Turris Omnia). However after about a week, the light turns off and the unit cannot be contacted, nor is it transmitting a signal. Unplugging and replugging makes it work again. I’ve got a POE module connected but am using the included wall wart. Hoping this is not a hardware problem, but simply some weird sleep condition…

1 Like

Maybe you have the same problem as me (and others) of the device rebooting and then being stalled? Did you check what happens when you try to reboot it?

Rebooting by pulling the power works every time. The logs don’t show anything unusual. There’s no software access of course so no other way to reboot, as far as I know.

rebooting by pulling the power works for me too, but its not ideal I think. especially when you don’t know what MOX is doing when he stops responding. it can make further issue with SD card etc.

Oh, now I get it! I rebooted while it was operational using the software reboot, and sure enough, it craps out. So perhaps it’s rebooting on its own and then stalling out. What’s the solution?

That’s exactly what I meant. It’s a known issue and apparently under investigation by Turris support. I have the same problem and so have several others. It seems that some people don’t have the problem. So there’s no solution known yet, unfortunately.

Thanks, good to know. Is there another thread here that I can follow where this has been discussed and we can expect a solution eventually?

Yes - here Reboot doesn't work

and check also https://gitlab.labs.nic.cz/turris/turris-build/issues/82

and if you are brave look at

Noticed that this problem got fixed, at least on my device, and so wanted to document that here. Presumably that was the result of the last commit…

@ mostlyharmlessone Thanks for the info. Which branch are you using? I tried yesterday, switched to HBT branch, but reboot still doesn’t seem to work for me.

Hmm, spoke too soon. My device doesn’t randomly freeze any more and runs successfully for weeks, but software reboot still hangs. So I’m not sure what was causing the previous problem I had. I had assumed it was trying to reboot. Anyway, I’m on TurrisOS 4.0.1 6dd7d54 / LuCI branch (git-19.281.83314-e07781f)

@Pepe do you know if this is meanwhile fixed in one of the development branches?
I noted that in the GitLab, @mbehun wrote about a month ago he worked on it. But is it fixed?
( https://gitlab.labs.nic.cz/turris/mox-boot-builder/commit/767f4c7ee11050883d1668ebc9c4156a4725e1d0 )

We are still investigating this issue and working on it. If there is anything new, we will let you know about it here and also in the Gitlab issue.

2 Likes

Hi @Pepe any news on this? Is it the work done by Marek 2 months ago part of the latest release? Is it supposed to fix the “stuck on reboot” issue? ( I am feeling very uncomfortable with my home potentially becoming unconnected until I go home and manually pull/push the power plug… )

Dear Turris users,

the Mox sometimes stuck on reboot issue is of the highest priority for the kernel team.

Since the problem is unpredictable when using the standard SOC reset, and the SOC can then be stuck in unusable state, we are solving this by doing the reset via software from the secure firmware instead (this means that the secure firmware puts the main CPUs and all peripherals into reset and then out of it).

The status is that this approach works most of the time, but sometimes after reboot the newly booted system is stuck - there are some issues regarding low level CPU settings. I am working on this and am confident that I will succeed soon. (But currently the probability of a successful reboot is still better when using the SOC reset.)

3 Likes

Thanks for that update! Actually for my MOX it’s not “sometimes stuck on reboot” but “always stuck on reboot” so I would be interested to know if there’s a way to reboot (command line?) with your new approach?

@Pepe Just wondering, is this fixed with the latest updates?

The reason why I’m asking is that I have the Turris in my remote house and if I reboot it now remotely and this is not fixed, then I will no longer be able to access it (and have to drive 4 hours to do so).
So if any doubt, I prefer not to reboot (but then I cannot update).
Can you let me know how I can check remotely/command line if it is fixed without rebooting?
With the latest update it didn’t work for me, I still had to pull the power cord.
I am now using TurrisOS 5.0.2, Linux turris 4.14.180 #0 SMP Sun Jun 14 23:12:16 2020 aarch64 GNU/Linux

For now, it is not fixed in the stable releases and that’s why it is mentioned in the known bugs. If it is going to be fixed, we will let you know. And looking into the mentioned pull request, it is just a hype. Nothing else. Milestone can be moved/removed and so on. I think you should rather wait until it is announced by us.

Thanks! I was indeed wondering given that there seems to be something in the github where that fix was included…

If not in stable - is the fix already in one of the “experimental” solutions?

1 Like