LSI megasas and duplicate erroneous devices?

I'm seeing the strangest thing with my M1015/LSI card.

I'm getting regular DID_BAD_TARGET errors in the kernel log for a drive that was erroneously detected.

Inital detection looks like:

[    0.860853] megasas: 06.806.08.00-rc1
[    0.860877] megasas: 0x1000:0x0073:0x1014:0x03b1: bus 1:slot 0:func 0
[    0.861172] megasas: FW now in Ready state
[    0.861796] libata version 3.00 loaded.
[    0.907832] megasas_init_mfi: fw_support_ieee=67108864
[    0.907902] megasas: INIT adapter done
[    0.955875] megaraid_sas 0000:01:00.0: Controller type: iMR
[    0.955990] scsi host0: LSI SAS based MegaRAID driver
[    0.959560] scsi 0:0:9:0: Direct-Access     ATA      ST3000DM001-1CH1 CC29 PQ: 0 ANSI: 5
[    0.964976] scsi 0:0:10:0: Direct-Access     ATA      ST3000DM001-1CH1 CC29 PQ: 0 ANSI: 5
[    0.970747] scsi 0:0:11:0: Direct-Access     ATA      ST3000DM001-1CH1 CC29 PQ: 0 ANSI: 5
[    0.975890] scsi 0:0:13:0: Direct-Access     ATA      WDC WD30EFRX-68E 0A82 PQ: 0 ANSI: 5
[    0.981458] scsi 0:0:15:0: Direct-Access     ATA      ST3000DM001-1ER1 CC25 PQ: 0 ANSI: 5
[    0.987097] scsi 0:0:16:0: Direct-Access     ATA      WDC WD30EFRX-68E 0A82 PQ: 0 ANSI: 5

Then after a while, the following appears:

[ 2545.701262] scsi 0:0:14:0: Direct-Access     ATA      WDC WD30EFRX-68E 0A82 PQ: 0 ANSI: 5

Note, this drive is a duplicate of either 0:0:13:0, or 0:0:16:0, there are only two WD Reds in this system. Two of the ports on the card are unpopulated.

Then I see some errors some time later:

[ 7113.505094] sd 0:0:14:0: [sdi] FAILED Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
[ 7113.506044] sd 0:0:14:0: [sdi] CDB: Read(16) 88 00 00 00 00 00 00 00 00 00 00 00 00 08 00 00
[ 7113.506967] blk_update_request: I/O error, dev sdi, sector 0
[ 7113.508415] sd 0:0:14:0: [sdi] FAILED Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
[ 7113.509330] sd 0:0:14:0: [sdi] CDB: Read(16) 88 00 00 00 00 01 5d 50 a3 00 00 00 00 08 00 00
[ 7113.510252] blk_update_request: I/O error, dev sdi, sector 5860532992
[ 7113.511567] sd 0:0:14:0: [sdi] FAILED Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
[ 7113.512522] sd 0:0:14:0: [sdi] CDB: Read(16) 88 00 00 00 00 01 5d 50 a3 a0 00 00 00 08 00 00
[ 7113.513487] blk_update_request: I/O error, dev sdi, sector 5860533152
[ 7113.515208] sd 0:0:14:0: [sdi] Synchronizing SCSI cache

This keeps happening over and over.

Attempting to smartctl -a on sdi fails with "no such device", and sdi does not currently appear in /dev

This machine is currently running a 4.0.2-1 kernel from debian sid.

What exactly can cause this, and how can I fix it?

Thomas Fjellstrom
[email protected]

[snip]

*sigh* It appears there are actually three WD Reds in the system. So the problem isn't near as weird as i thought. I noticed the drive was actually visible a few minutes ago and it had a different S/N than the other WD's. I decided to take a look at the drives, and make sure all cables were seated properly, and while i was in there i took a can of compressed air to the entire box. Upon boot the drive was detected immediately, and so far hasn't dropped off, but it never dropped right away, so we'll see what happens. I'm running badblocks -w on it to see if it encounters any errors at all. If theres no errors, It'll go back into service as a spare in a raid6 array.

I've had to mess with that system so many times the past year I forgot I ended up putting three new WD's in there after a few Seagate's failed.

So there we go. Nothing here unless it drops out again. I apologize for the noise.

[snip]

Thomas Fjellstrom
[email protected]