ZyXEL NSA320 booted from USB...once!

[WarheadsSE]

Well that figures. I may have to dig out the deep magics if stuff like that wants to keep happening.

[ormiller]

New Problem.
I did a pacman -Su and updated the system. When I was done I typed in poweroff. In the morning when I powered the NSA325 it would not boot. I pulled the drive and placed it on a usb drive mate connected to my laptop with linux.
(By the way the update did the same to my Pogoplug Series 4, and the superblock on usb stick is trashed also.)
Here is the output from the fsck

# fsck /dev/sdd1
fsck from util-linux-ng 2.17.2
e2fsck 1.41.12 (17-May-2010)
fsck.ext2: Superblock invalid, trying backup blocks...
fsck.ext2: Bad magic number in super-block while trying to open /dev/sdd1

The superblock could not be read or does not describe a correct ext2
filesystem. If the device is valid and it really contains an ext2
filesystem (and not swap or ufs or something else), then the superblock
is corrupt, and you might try running e2fsck with an alternate superblock:
e2fsck -b 8193 <device>

_________________________
For my Zyxel I took another hard drive set it up like before ran the script with the modifications for ext2.
Now my box restarts every 32 seconds. Archlinux will get up and running. Now what ever it is, or is not doing it restarts every 32 seconds. If I stop the boot process at the prompt and let it sit doing nothing it will still recycle at the 32 second mark. When makes it to the alarm prompt when it hits the 32 seconds mark no matter what it is doing it restarts.

Is there an environment variable that gets set that can cause it to recycle every 32 seconds ?

I did a run to_stock and it does not have the recycle problem.

[WarheadsSE]

Sounds like an issue with the watchdog

[CharlesB]

$this->bbcode_second_pass_quote('WarheadsSE', 'S')ounds like an issue with the watchdog

[WarheadsSE]

There is an orion watchdog module, we'll look into what was happening as soon as we get a chance.

[WarheadsSE]

Alright, on the 320:
We're still hunting this thing down, and getting it to misbehave for me is getting to be really annoying. I have to leave it powered off for 4 days this last time.

What we have found along the way, is merely that when linux has no connection, neither does uboot. So clearly, it's something neither of them is doing.

We have also learned how to set the return from power loss behavior, which I will share below.

$this->bbcode_second_pass_code('', '
#!/bin/bash
# power_resume.sh [0/1]
# 0 to disable, 1 to enable.
# enable/disable resume from power loss.

# export both gpios
echo 47 > /sys/class/gpio/export
echo 49 > /sys/class/gpio/export

# 49 out, value 1
echo "out" > /sys/class/gpio/gpio49/direction
echo 1 > /sys/class/gpio/gpio49/value
# 47 out, value 1
echo "out" > /sys/class/gpio/gpio47/direction
echo $1 > /sys/class/gpio/gpio47/value
# set 49 value 0
echo 0 > /sys/class/gpio/gpio49/value
')

[WarheadsSE]

Hah. Cheater method.

Apparently, the PHY eth stays happy if you turn resume ON. So, use that script to set it on (./power_resume.sh 1)

[CharlesB]

Since I have a 2nd device (which I hadn't touched yet) I figured I'd retry the USB install method (this time with the smaller drive that I know works w/ uboot) on the new device, and see if its Ethernet port behaves more like my other one (consistently bad after power-up) or like WarheadsSE's (occasionally bad).

Of course, now that I have the Bus Pirate, I wanted to see what I didn't get to see before. Glad I did; the USB installer still didn't work right, but I got a capture that might be useful at some point.

Updating the NAND went fine, but the installation to the hard drive seemed to have problems.

Here's the capture. I placed a few notes in it that look like this:$this->bbcode_second_pass_quote('', '(')(( This is a note )))

[WarheadsSE]

I can actually repeatable cause it now. The OOM killer however, doesn't make sense why it is triggering, but it all seems to be down to that happening. The uImage load is down to what appears to be a fault with writing to the disk or a memory issue.

[ormiller]

In response to my possible watchdog error reboot with the NSA325.

I placed a small unformatted drive into bay 1. Went through the Zyxel initialization process. Rebooted the box. The system ran for days. From the serial console I issued a reboot. At the MARVELL> prompt I issued a run to_linux. I quickly pulled drive 1 then quickly inserted the ArchLinux drive and now I have a functioning system once again.

when my system was rebooting every 23 seconds. I did notice once during a reboot that there was is a message

sched_clock: 32 bits at 200MHz, resolution 5ns, wraps every 21474ms

which is real close to my 23 seconds

If I issue a # poweroff then I get the 23 second reboot again and have to go through the run to_stock insert the Zyxel formatted disk. Wait till it is up and running, then reboot then run to_linux insert the ArchLinux disk to be up and running.