NSA320 Fails to start system logger

[lancej29]

I just got a ZyXEL NSA320 and tried to install Archlinux using the USB stick (usb_key_func) technique given on the Archlinux NSA320 page. I have connected to its serial port. My NSA320 has V4.50 firmware which, apparently, no longer has mkfs.ext3 so the script fails.

It does, however, have mkfs.ext4, so I changed the ext3 references to ext4 in the script and completed the installation. The unit booted the kernel, mounted the root file system and began initialization. During initialization it started repeatedly showing the following:

$this->bbcode_second_pass_code('', '
[FAILED] Failed to start System Logger Daemon.
See 'systemctl status syslog-ng.service' for details.
Starting System Logger Daemon...
')

The messages came so fast, I had to power the unit down so I could read them. Mixed in with those messages were some messages like:

$this->bbcode_second_pass_code('', '
[FAILED 114.267299] systemd[1]: syslog-ng.service start request repeated too quickly, refusing to start.
m] Failed to start System Logger Daemon.
See 'systemctl status syslog-ng.service' for details.
Starting System Logger Daemon...
')

Here are the last few OK messages just before the failures started:

$this->bbcode_second_pass_code('', '
[ OK ] Started Periodic Command Scheduler.
Starting Netcfg multi-profile daemon...
Starting SSH Key Generation...
Starting Login Service...
Starting D-Bus System Message Bus...
[ OK ] Started D-Bus System Message Bus.
Starting System Logger Daemon...
[ OK ] Started System Logger Daemon.
[ 108.371185] systemd-journald[73]: Received SIGUSR1
[ OK ] Started Trigger Flushing of Journal to Persistent Storage.
Starting Permit User Sessions...
[ OK ] Started Login Service.
[ OK ] Started Permit User Sessions.
Starting Getty on tty1...
[ OK ] Started Getty on tty1.
Starting Serial Getty on ttyS0...
[ OK ] Started Serial Getty on ttyS0.
[ OK ] Reached target Login Prompts.
Starting System Logger Daemon...
[ OK ] Started System Logger Daemon.
[ 109.207915] ADDRCONF(NETDEV_UP): eth0: link is not ready
Starting System Logger Daemon...
[ OK ] Started System Logger Daemon.
Starting System Logger Daemon...
[ OK ] Started System Logger Daemon.
Starting System Logger Daemon...
[ OK ] Started System Logger Daemon.
Starting System Logger Daemon...
[ OK ] Started System Logger Daemon.
')

Any ideas what is going on? What additional information will you need?

Thanks in advance,

Lance

[WarheadsSE]

Irk, darn commands coming and going. Thanks for the heads up on that one. The wonk logger, check that that rootfs is being mounted rw.

[lancej29]

Warheads, thanks for the quick response. I don't really know how to check if the rootfs is being mounted ro or rw since I can't do anything. The system won't respond to ssh or telnet so I only have the serial console which is flooded with messages. The box won't respond to pongs either. (I know the IP address from the router.) I can see a login prompt scroll by if I hit enter a lot. Any suggestions?

[WarheadsSE]

catch the output of "kernel command line: "

[lancej29]

The kernel command line had the file system mounted read only. That is because it was in the u-boot bootargs that way, which in turn, was because the "fw_setenv bootargs_linux" line in usb_key_func.sh.2 had it that way. I didn't think to look there.

I assume that this is intentional and that somewhere in the boot/init process the rootfs is supposed to be remounted rw. Any idea where this might be breaking down?

I changed the bootargs to make the file system rw and it booted without failures. However, I did see these ominous messages:

$this->bbcode_second_pass_code('', '
[ 17.601606] ata2: SATA link down (SStatus 0 SControl F300)
[ 17.633735] EXT4-fs (sda2): couldn't mount as ext3 due to feature incompatibilities
[ 17.651858] EXT4-fs (sda2): couldn't mount as ext2 due to feature incompatibilities
[ 19.088319] EXT4-fs (sda2): warning: checktime reached, running e2fsck is recommended
[ 19.545755] EXT4-fs (sda2): recovery complete
[ 19.550469] EXT4-fs (sda2): mounted filesystem with ordered data mode. Opts: (null)
[ 19.558218] VFS: Mounted root (ext4 filesystem) on device 8:2.
[ 19.564170] devtmpfs: mounted
')

Does the init process try to mount /dev/sda2 as ext3 somewhere along the line? (I am not familiar with arch boot/init process.)

Sorry to be pestering you on Christmas day, but it's one of the few times that I get time do such stuff. If you find time to answer, that's great, but it's not urgent. Please enjoy your holiday, and thanks for your help.

[cyberic]

It seems that everything is fine but the login process waits for a console...
Do you have something like 'console=ttyS0' in your bootloader env ?

[lancej29]

Actually. by making the root file system rw in the bootargs, the system booted and I can log in both at the serial console and via ssh.

My concern at this point is why the bootargs set by the Archlinux USB stick had the root file system mounted as ro. It's hard to believe that this was a mistake because everyone would be having the problem (although everyone else might have been smart enough to figure it out by themselves). So I am wondering if the file system was supposed to be changed to rw somewhere in the init process, but that might have failed.

[cyberic]

I think that systemd is supposed to remount it rw

[WarheadsSE]

I think it is a lantent hold over from old initscripts setup. I have to update it to hanfle the ext3/4 switch out anyways!

[planeteater]

$this->bbcode_second_pass_quote('cyberic', 'I') think that systemd is supposed to remount it rw