I have a small home server built around a Lian Li PC-Q25 case with a SATA backplate advertising itself as hot pluggable. The motherboard is Asus P8H77-I. I have 4 SATA drives attached to the backplate - two pairs built into RAID1 arrays. The system is running on Centos 6.3 x86_64.
One of the drives broke down, so I did the recommended procedure: synced, removed it from the array, shut it down properly and pulled it out. No disaster here, I could hear the drive spin down and no errors appeared in the dmesg log.
Now, I assumed that by the SATA standard the staggered pins in the drive would ensure a safe plug-in without any sudden power surge. Pushing the drive in, I could head the other drives slow down and click their heads for a very brief moment.
Checking the dmesg log revealed the following:
ata1: exception Emask 0x10 SAct 0x0 SErr 0x4090000 action 0xe frozen
ata3.00: exception Emask 0x10 SAct 0x3ff007ff SErr 0x4890000 action 0xe frozen
ata3.00: irq_stat 0x08400040, interface fatal error, connection status changed
ata3: SError: { PHYRdyChg 10B8B LinkSeq DevExch }
ata3.00: failed command: WRITE FPDMA QUEUED
ata3.00: cmd 61/80:00:3f:81:ca/00:00:00:00:00/40 tag 0 ncq 65536 out
res 40/00:54:bf:81:ca/00:00:00:00:00/40 Emask 0x10 (ATA bus error)
ata3.00: status: { DRDY }
(The last ata3.00 messages are repeated about 20 times with different numbers but the same text)
The last lines are:
ata3.00: status: { DRDY }
ata3: hard resetting link
ata2: exception Emask 0x50 SAct 0x0 SErr 0x4090800 action 0xe frozen
ata2: irq_stat 0x00400040, connection status changed
ata2: SError: { HostInt PHYRdyChg 10B8B DevExch }
ata2: hard resetting link
ata1: irq_stat 0x00400040, connection status changed
ata1: SError: { PHYRdyChg 10B8B DevExch }
ata1: hard resetting link
ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
ata1.00: configured for UDMA/133
ata1: EH complete
ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata2.00: configured for UDMA/133
ata2: EH complete
ata3.00: configured for UDMA/133
ata3: EH complete
Also my logwatch reported the following changes in SMART data:
/dev/disk/by-path/pci-0000:00:1f.2-scsi-2:0:0:0 [SAT] :
Prefailure: Raw_Read_Error_Rate (1) changed to
100,
Prefailure: Reallocated_Sector_Ct (5) changed to
200,
Prefailure: Spin_Up_Time (3) changed to
100,
Usage: Seek_Error_Rate (7) changed to
200,
/dev/disk/by-path/pci-0000:00:1f.2-scsi-3:0:0:0 [SAT] :
Usage: Calibration_Retry_Count (11) changed to
100,
Usage: Load_Retry_Count (223) changed to
100,
Device: /dev/disk/by-path/pci-0000:00:1f.2-scsi-2:0:0:0 [SAT], Self-Test Log error count increased from 0 to 1
On the following day the SMART log still had suspicious entries in it:
/dev/disk/by-path/pci-0000:00:1f.2-scsi-0:0:0:0 [SAT] :
Usage: Seek_Error_Rate (7) changed to
200,
/dev/disk/by-path/pci-0000:00:1f.2-scsi-1:0:0:0 [SAT] :
Usage: Seek_Error_Rate (7) changed to
200,
/dev/disk/by-path/pci-0000:00:1f.2-scsi-2:0:0:0 [SAT] :
Usage: Multi_Zone_Error_Rate (200) changed to
200,
/dev/disk/by-path/pci-0000:00:1f.2-scsi-3:0:0:0 [SAT] :
Usage: Throughput_Performance (2) changed to
56,
/dev/disk/by-path/pci-0000:00:1f.2-scsi-4:0:0:0 [SAT] :
Prefailure: Raw_Read_Error_Rate (1) changed to
116, 117,
Usage: ECC_Uncorr_Error_Count (195) changed to
116, 117,
So, apparently the sata backend just brutely powered the drive on immediately, possibly causing the voltage to drop for a moment.
My mistake was perhaps to plug all four drives in the same PSU rail and expect the PSU (albeit a 800W Seasonic with good specifications) to cope with the sudden power draw.
The SATA backplate has two Molex connectors on the back for power - I'll plug them to separate PSU rails to ensure a steadier power output.
Is there a way to prevent the drive from spinning up immediately as I stick it back into the drive pack?
Also, did I possibly just damage the drives (can it be seen from these log messages)?
Thank you!