NWSP1 / Temp Guard rebooting loop

[Jocko]

so far so good. there is your collaborated issue with the standby binary.
Will you fix the code in an subsequent firmware release or update?

With Fvdw, we must understand what is wrong. I am pretty sure that it ran well with the previous firmware versions. :scratch

the second issue is that (even when tempguard is disabled) the restart buttom
will not do a restart; the box won,t come up again.

can you reproduce that too? or does that behavior also relate to the standby binary?

many thank and I say again, congrats the cool work you have done..

greetings marcel

Strange because I have no problem with the button. I can stop and start my NAS with it if you follow the sequence as described previously. Same behavior, if I use the button in the web-interface :?

[Jocko]

Some progress but not good news... :crazy

It is again because the nwsp1 has not a push button. :evil:

In a shutdown sequence, if the power button position is on 1 then the nas restarts logically (of course) !!!! :doh

but if the position is 0 then the NAS doesn't restart.

So a fix could be to set the power button on 0 after booting : for example just after the first blinking sequence. And in this case, the shutdown button runs well and the booting loop can't appear if you enable the temp-guard service. :)

But there is a new issue : the restart button can't run ! the NAS will stop but won't restart... :whistle

In summary, I don't think there is a fix what will allow to keep all the features available in the same time : stop/restart/wol

[Mijzelf]

If you can 'wol' now, you probably have a way to store some flag over a reboot, right? So you could at least bring it to 'wol' on a temp-guard shutdown. If you have room for an extra flag, you could put the kernel in a low-power state without enabling the network port. That should bring the power consumption down to less than 2W. That should be enough as an adequate respons on a temperature-too-high status.

[fvdw]

If you can 'wol' now, you probably have a way to store some flag over a reboot, right? So you could at least bring it to 'wol' on a temp-guard shutdown. If you have room for an extra flag, you could put the kernel in a low-power state without enabling the network port. That should bring the power consumption down to less than 2W. That should be enough as an adequate response on a temperature-too-high status.

Good suggestion, we could do that. PS the lacie "wol" is not real wol the power on the network interface must be on to detect the magic packet. the only difference with active is that the HDD is spin down and there run no programs only the program checking for the magic packet. (this even the case in the more sophisticated lacie's it fakes WOL)