Re: Wheezy boot stops, leaving machine unreachable

[David Christensen <dpchrist@holgerdanske.com>]

On 01/11/2015 04:07 PM, Gary Dale wrote:
> I've been a experiencing this problem for a couple of months
...
> > [    7.672023] ata2.00: qc timeout (cmd 0xec)
> > [    7.672061] ata2.00: failed to IDENTIFY (I/O error, err_mask=0x4)
> > [   13.204015] ata2: link is slow to respond, please be patient (ready=0)
> > [   17.684019] ata2: softreset failed (device not ready)
> > [   18.176021] ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
> > [   28.176046] ata2.00: qc timeout (cmd 0xec)
> > [   28.176084] ata2.00: failed to IDENTIFY (I/O error, err_mask=0x4)
> > [   28.176115] ata2: limiting SATA link speed to 3.0 Gbps
> > [   28.668019] ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
> > [   28.668527] ata2.00: ATA-9: WDC WD2003FZEX-00Z4SA0, 01.01A01, max
> > UDMA/133
> > [   28.668558] ata2.00: 3907029168 sectors, multi 0: LBA48 NCQ (depth
> > 31/32), AA
> > [   28.669061] ata2.00: configured for UDMA/133

Linux is having trouble connecting to the WD Black hard drive.


You might have an bad cable and/or dirty connection.

- Power-off your machine, leave the power cable connected (e.g. 
protective earth connected), open the case, and vacuum it out.

- Unplug the SATA cable verify it's rated for 6 Gbps.  If not, get one 
that is.

- Plug in a rated cable.

- Swap cable with known good, rated cable.


Verify that /etc/crypttab, /etc/fstab, etc., use UUID's and/or id's, and 
then:

- Try using a different controller and/or port.

- If you're using an expansion card (HBA), unplug and replug it.  Swap 
it if you have a spare.


> > [   57.647787] SAK: killed process 1943 (mount):
> > task_session(p)==tty->session
> > [   57.647787] SAK: killed process 1944 (mount.cifs):
> > task_session(p)==tty->session
> > [   57.647787] SAK: killed process 1945 (mount.cifs):
> > task_session(p)==tty->session

SAK is related to your serial port:

http://unix.stackexchange.com/questions/80857/my-process-was-killed-but-i-cannot-understand-the-kernel-notice


Pay attention to your power-on/ power-off sequences.


Test the serial cable.  Test the serial port.  Test the serial adapter.


> > [  283.660852] ata2.00: irq_stat 0x08000000, interface fatal error
> > [  283.664407] ata2: SError: { UnrecovData 10B8B BadCRC }
> > [  283.667944] ata2.00: failed command: READ FPDMA QUEUED
> > [  283.671490] ata2.00: cmd 60/00:d8:00:80:90/04:00:01:00:00/40 tag 27
> > ncq 524288 in
> > [  283.671493]          res 40/00:dc:00:80:90/00:00:01:00:00/40 Emask
> > 0x10 (ATA bus error)
> > [  283.678625] ata2.00: status: { DRDY }
> > [  283.682172] ata2: hard resetting link
> > [  284.176193] ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
> > [  284.181280] ata2.00: configured for UDMA/133
> > [  284.185347] ata2: EH complete

There goes the WD Black again.


You might have two issues -- the WD Black and the serial port.


You will likely find it useful to save the output of dmesg as you 
troubleshoot.  Each time you boot, save it to a file with a meaningful 
name.  It might help to also save a copy without the timestamps:

    $ dmesg > testname.out

    $ dmesg | perl -ne 'm/\[.+?\] (.+)/; print qq($1\n)' > testname.out2


Keep a running log in a text file as you go.  Cut and paste console 
sessions verbatim.


Run tests multiple times to confirm results are repeatable.


HTH,

David