Hi,

I recently bought an intel SAS expander RES2SV240, this is paired with IBM M1015 flashed to IT mode and firmware P19 (read few posts about P20 having issues). I am using both the SAS ports from the IBM as inputs to the Intel expander. All the disks are recognized and seems to be okay from outside. I started a parity check to see how the data transfer speeds compare without the Intel expander. It as been only few hours since I started the parity check and as of now, the speeds 'appear' to be only slightly slower. I was watching a video simultaneously on my Win8.1 KVM and noticed the video 'slightly' stutter. Checking the logs, I see some strange warnings and errors that I haven't seen before. Shown below is a snippet and I have attached the entire log. I am not sure what 'ata6' is being referred here, and whether the errors point to a lose sata cable on one disk or a sas cable or something wrong with the expander itself?

 

Almost all the errors seem to have occurred around 16:00 hrs.

 

On another note, is there a way to get a continuous record of the parity check speeds, time or % complete vs check speed in MB/s? It would be great to compare two scenarios with all the disks spinning and accessing data.

 

Thanks!

 

...

Dec 18 16:00:02 Tower kernel: ata6: EH complete

Dec 18 16:00:02 Tower kernel: ata6.00: exception Emask 0x10 SAct 0x0 SErr 0x400100 action 0x6 frozen

Dec 18 16:00:02 Tower kernel: ata6.00: irq_stat 0x08000000, interface fatal error

Dec 18 16:00:02 Tower kernel: ata6: SError: { UnrecovData Handshk }

Dec 18 16:00:02 Tower kernel: ata6.00: failed command: WRITE DMA EXT

Dec 18 16:00:02 Tower kernel: ata6.00: cmd 35/00:00:18:77:00/00:04:00:00:00/e0 tag 7 dma 524288 out

Dec 18 16:00:02 Tower kernel: ata6.00: status: { DRDY }

Dec 18 16:00:02 Tower kernel: ata6: hard resetting link

Dec 18 16:00:02 Tower kernel: ata6: SATA link up 6.0 Gbps (SStatus 133 SControl 300)

Dec 18 16:00:02 Tower kernel: ata6.00: ACPI cmd ef/10:06:00:00:00:00 (SET FEATURES) succeeded

Dec 18 16:00:02 Tower kernel: ata6.00: ACPI cmd f5/00:00:00:00:00:00 (SECURITY FREEZE LOCK) filtered out

Dec 18 16:00:02 Tower kernel: ata6.00: ACPI cmd b1/c1:00:00:00:00:00 (DEVICE CONFIGURATION OVERLAY) filtered out

Dec 18 16:00:02 Tower kernel: ata6.00: ACPI cmd ef/10:06:00:00:00:00 (SET FEATURES) succeeded

Dec 18 16:00:02 Tower kernel: ata6.00: ACPI cmd f5/00:00:00:00:00:00 (SECURITY FREEZE LOCK) filtered out

Dec 18 16:00:02 Tower kernel: ata6.00: ACPI cmd b1/c1:00:00:00:00:00 (DEVICE CONFIGURATION OVERLAY) filtered out

Dec 18 16:00:02 Tower kernel: ata6.00: configured for UDMA/133

.....

Logs_WithSASExpander.zip

In this case, ata6 corresponds to the sixth motherboard SATA port, your Parity drive, sdg, a Seagate 4tb with serial ending in Z18.  The problem only lasted a few seconds, and quit once it dropped the SATA link speed to 3.0 Gbps (from 6.0 Gbps), appears to be otherwise harmless.  The Parity drive is driven harder than other drives during certain operations, so it may have been pushing a little too fast for that port, even if it claimed to be able to handle 6.0 Gbps.  You might try swapping the first and 6th drives, try the parity drive on the very first SATA port.  Otherwise, I can't tell for sure what failed, other than it was a communications failure somewhere in the interface to the drive, and once throttled, worked fine.  It will probably do that again, when it feels the need to.

 

One oddity, the very first exception reported both the RecovData and UnrecovData flags, opposites of each other.  One says it recovered, the other says it couldn't!  Never seen that happen before.

 

The SAS expander was not involved at all.