[GTALUG] Remove drive, system won't boot

[Scott Sullivan]

On 1/4/20 11:24 AM, William Park via talk wrote:
> Looks like your HDD1 (Seagate 4TB) is failing, so can't mount /var,
> /home, etc.  But, the actual mounting order goes like
>      sda -- Adata 60GB
>      sdb -- WD 10TB (the new disk)
>      sdc -- Seagate 4TB
> 
> 1. Try swapping the SATA connectors of 10TB and 4TB disks.
> 2. Replace 4TB as well.
> -- William Park <opengeometry at yahoo.ca> On Sat, Jan 04, 2020 at 

Evan,

You don't outline which HDDX is which. But as William points out, your 
4TB Seagate drive is not behaving well. Sometimes these errors can be a 
the result the sata port going bad, or a bad cable. But mostly it's the 
drive failing. Is that the drive with your /var partition? Because if it 
is, that is what's hanging your boot.


Jan 04 08:06:35 bluejay kernel: ata3.00: ATA-8: ST4000DM000-1F2168, 
CC52, max UDMA/133

Jan 04 08:06:35 bluejay kernel: ata3.00: exception Emask 0x10 SAct 
0x2000 SErr 0x4090000 action 0xe frozen
Jan 04 08:06:35 bluejay kernel: ata3.00: irq_stat 0x00400040, connection 
status changed
Jan 04 08:06:35 bluejay kernel: ata3: SError: { PHYRdyChg 10B8B DevExch }
Jan 04 08:06:35 bluejay kernel: ata3.00: failed command: READ FPDMA QUEUED
Jan 04 08:06:35 bluejay kernel: ata3.00: cmd 
60/00:68:00:b4:c0/01:00:d1:01:00/40 tag 13 ncq 131072 in
                                          res 
40/00:68:00:b4:c0/00:00:d1:01:00/40 Emask 0x10 (ATA bus error)
Jan 04 08:06:35 bluejay kernel: ata3.00: status: { DRDY }
Jan 04 08:06:35 bluejay kernel: ata3: hard resetting link

Jan 04 08:06:45 bluejay kernel: ata3: exception Emask 0x10 SAct 0x0 SErr 
0x4090000 action 0xe frozen
Jan 04 08:06:45 bluejay kernel: ata3: irq_stat 0x00400040, connection 
status changed
Jan 04 08:06:45 bluejay kernel: ata3: SError: { PHYRdyChg 10B8B DevExch }
Jan 04 08:06:45 bluejay kernel: ata3: hard resetting link
Jan 04 08:06:51 bluejay kernel: ata3: link is slow to respond, please be 
patient (ready=0)
Jan 04 08:06:51 bluejay kernel: ata3: SATA link up 3.0 Gbps (SStatus 123 
SControl 300)
Jan 04 08:06:51 bluejay kernel: ACPI Error: [DSSP] Namespace lookup 
failure, AE_NOT_FOUND (20150930/psargs-359)
Jan 04 08:06:51 bluejay kernel: ACPI Error: Method parse/execution 
failed [\_SB.PCI0.SAT0.SPT2._GTF] (Node ffff8802174d61e0), AE_NOT_FOUND 
(20150930/psparse-542)
Jan 04 08:06:51 bluejay kernel: ACPI Error: [DSSP] Namespace lookup 
failure, AE_NOT_FOUND (20150930/psargs-359)
Jan 04 08:06:51 bluejay kernel: ACPI Error: Method parse/execution 
failed [\_SB.PCI0.SAT0.SPT2._GTF] (Node ffff8802174d61e0), AE_NOT_FOUND 
(20150930/psparse-542)
Jan 04 08:06:51 bluejay kernel: ata3.00: configured for UDMA/133
Jan 04 08:06:51 bluejay kernel: ata3: EH complete

> 09:49:58AM -0500, Evan Leibovitch via talk wrote:
>> Hi again,
>>
>> Sorry for two system problems over the holidays, but this last one truly
>> has me stumped.
>>
>> Last week I had a problem trying to back up files onto a large drive using
>> a USB enclosure.
>>
>> This week I've tried to complete the install, and of course it could not go
>> without problems....
>>
>> The config is:
>> SDD with /boot and /
>> HDD1 with /var, /home and swap
>> HDD2 has one data partition
>> HDD3 has one data partition
>> I am trying to replace HDD2 with HDD3.
>>
>> HDD3 installs and mounts just fine. But if I unplug HDD2, the machine
>> refuses to boot. Taking it to recovery mode eventually ends up in a
>> hardware freeze. Plugging it back in goes to a normal boot, even though the
>> HDD2's partition is no longer mounted to anything.
>>
>> I've tried to eliminate the BIOS as a source of the problem, have updated
>> it to the latest version and ensured that the disk-to-be-removed is not
>> seen in any BIOS configurations.
>>
>> I attach the one log file I am able to capture, since it freezes before
>> mounting /var.
>>
>> As best as I can tell the freeze happens during the point of the bootup
>> where it's running fsck on the drives. It may be looking for the removed
>> HDD2 but I don't know where it's remembering to look for it. HDD2 is no
>> longer in /etc/fstab and I searched in vain for any instance of the UUID of
>> HDD2 in either /boot or /etc.
>>
>> Any suggestions?

-- 
Scott Sullivan