I have not done a parity check since April, so I decided to do one last night.  The system was showing a few hundred thousand parity errors, so I just let it run.  Hours later I cannot get to the web interface, and it is not pinging.  The drive activity light is no longer flashing.  I hit the reset button, but I do not have a monitor attached, so I do not know what happened.  I still cannot ping it.

 

Any ideas?

 

I will be attaching a monitor when I get home, so I can see what is going on.  I also just upgraded from 4.3.-beta6 to 4.3.2 a week ago.

If you can't ping it, it is unlikely you can telnet to it.  I am afraid you will probably need a monitor attached for troubleshooting in this case.

He's right, no substitute for eyes on the console.

 

I am aware of cases like this, where too many errors logged to the syslog filled memory, causing the kernel to close down processes.  So it's possible, that the console is still operational, if it's not too late.  You might try logging in blindly and capturing the syslog, before you reboot.  If system is still setup as in the default, try typing root, press the Enter key, then typing cp /var/log/syslog /boot.  (Probably) can't hurt, and might save the syslog for examination of why all the errors, and what was the very first error.

 

Side note: The size of the syslog, even if massively full of errors, should not overflow the memory, even if you only have 512MB, so I'm guessing the syslog entries must be kept in a very inefficient linked list, amplifying the memory requirements.

 

Edit:  Sorry, just reread your post, realized too late you had already hit the reset, so too late for that first syslog.

I connected a monitor and KB, and had to reboot again.

 

The system came up once and two of my drives said Mounting, but never did.  I rebooted one more time, and everything came back online.  The parity check cleared with zero errors.  Thanks for the help!