Hi

 

I had some problems a while back with disks going disabled and couldnt figure out why.

Some suggestions was changing the SATA cable, changing the psu etc

 

So what i did was

 

Changed PSU

Installed a 9201-16i HBA card

Bought new sff 8087 cables (some from ebay and some from a webshop)

Ebay

Webshop

 

All disks are connected to the HBA except the cache drive.

 

Now im getting these errors on two of the disks, one disk is unassigned and one is one of the parity drives

 

Mar 8 10:38:14 Tower kernel: print_req_error: I/O error, dev sdf, sector 0
Mar 8 10:38:14 Tower kernel: print_req_error: I/O error, dev sdf, sector 0
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 Sense Key : 0x2 [current]
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 ASC=0x4 ASCQ=0x0
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 CDB: opcode=0x8a 8a 00 00 00 00 00 86 51 67 a8 00 00 00 08 00 00
Mar 8 10:38:14 Tower kernel: print_req_error: I/O error, dev sdf, sector 2253481896
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 Sense Key : 0x2 [current]
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 ASC=0x4 ASCQ=0x0
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 CDB: opcode=0x8a 8a 00 00 00 00 00 ad 1d 9c 20 00 00 00 08 00 00
Mar 8 10:38:14 Tower kernel: print_req_error: I/O error, dev sdf, sector 2904398880
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 Sense Key : 0x2 [current]
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 ASC=0x4 ASCQ=0x0
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 CDB: opcode=0x8a 8a 00 00 00 00 00 a4 db ea 10 00 00 00 08 00 00
Mar 8 10:38:14 Tower kernel: print_req_error: I/O error, dev sdf, sector 2765875728
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 Sense Key : 0x2 [current]
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 ASC=0x4 ASCQ=0x0
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 CDB: opcode=0x8a 8a 00 00 00 00 03 57 0c 94 d0 00 00 00 08 00 00
Mar 8 10:38:14 Tower kernel: print_req_error: I/O error, dev sdf, sector 14345344208
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 Sense Key : 0x2 [current]
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 ASC=0x4 ASCQ=0x0
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 CDB: opcode=0x8a 8a 00 00 00 00 03 57 0c 95 c0 00 00 00 08 00 00
Mar 8 10:38:14 Tower kernel: print_req_error: I/O error, dev sdf, sector 14345344448
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 Sense Key : 0x2 [current]
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 ASC=0x4 ASCQ=0x0
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 CDB: opcode=0x8a 8a 00 00 00 00 03 57 0c 95 e0 00 00 00 08 00 00
Mar 8 10:38:14 Tower kernel: print_req_error: I/O error, dev sdf, sector 14345344480
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 Sense Key : 0x2 [current]
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 ASC=0x4 ASCQ=0x0
Mar 8 10:38:14 Tower kernel: sd 13:0:3:0: [sdf] tag#0 CDB: opcode=0x8a 8a 00 00 00 00 03 57 0c 96 00 00 00 00 08 00 00
Mar 8 10:38:52 Tower kernel: sd 13:0:3:0: [sdf] Synchronizing SCSI cache
Mar 8 10:38:52 Tower kernel: sd 13:0:3:0: [sdf] Synchronize Cache(10) failed: Result: hostbyte=0x01 driverbyte=0x00

 

Bought 2 more SFF 8087 cables from the local webshop today since these are perhaps higher quality than the ebay ones, will be here in couple of days

Anyone got any idea what could be wrong this time ?

tower-diagnostics-20180308-2051.zip

Make sure the cables are tight.How are you powering the drives? With adaptors or are they on a backplane? I just had a similar problem after upgrading my disks and cpu. In my case, somehow my molex to sata adaptor became flaky.

The disk in question is Powered straight from the psu With no adapters.

 

my data disks are on the Lian li hot swap backplanes (bp3sata), more incoming so all the disks are connected by them

Parity2 dropped offline so there's no SMART, but assuming it's OK like the unassigned disk, and since both share the same minSAS cable/port, I would start by changing those, you have an unused miniSAS port so swap it there, if more issues replace the cable when it arrives.

Forgot to mention, increase your cache floor setting, you let the cache fill up, and this will cause problems for the docker image:

 

Mar  6 15:33:04 Tower shfs: cache disk full
Mar  6 15:33:36 Tower shfs: cache disk full
Mar  6 15:33:36 Tower shfs: cache disk full
Mar  6 15:33:50 Tower kernel: loop: Write error at byte offset 531709952, length 4096.
Mar  6 15:33:50 Tower kernel: print_req_error: I/O error, dev loop2, sector 1038496
Mar  6 15:33:50 Tower kernel: BTRFS error (device loop2): bdev /dev/loop2 errs: wr 1, rd 0, flush 0, corrupt 0, gen 0
Mar  6 15:33:50 Tower kernel: loop: Write error at byte offset 531759104, length 4096.
Mar  6 15:33:50 Tower kernel: print_req_error: I/O error, dev loop2, sector 1038592

 

Thanks for the replies

 

gonna try it tonight

Changed the breakout cable and tried to plug it into different sas port.

No disks from that cable found at all.

 

Plugged it back into the previous sas port and presto, all disks detected.

 

Guess i have to rma the 9201-16i card.....