On 2003.02.10 10:41, Christoph Kaegi wrote:
> Dear NetBSDers
> 
> I set up RaidFrame mirroring last week one of our 
> Supermicro 6012P-6B. The have an Adaptec aic7899
> chipset onboard with two identical Seagate ST336607LC
> disks on it.
> 

OK, it died again tonight. This time it also trashed the
mirror and the softdeps enabled /usr filesystem.

-------------------------------------- 8< --------------------------------------
Feb 12 03:15:46 mx2 /netbsd: sd1(ahc0:0:2:0):  Check Condition on CDB: 0x2a 00 00 4d 44 4f 00 00 10 00
Feb 12 03:17:10 mx2 /netbsd:     SENSE KEY:  Aborted Command
Feb 12 03:17:10 mx2 /netbsd:      ASC/ASCQ:  Overlapped Commands Attempted
Feb 12 03:17:10 mx2 /netbsd:      FRU CODE:  0x1
Feb 12 03:17:10 mx2 /netbsd: 
Feb 12 03:17:10 mx2 /netbsd: sd0(ahc0:0:0:0): SCB 3a - timed out while idle, SEQADDR == 0xa
Feb 12 03:17:10 mx2 /netbsd: SCSIRATE == 0x0
Feb 12 03:17:10 mx2 /netbsd: sd0(ahc0:0:0:0): Queuing a BDR SCB
Feb 12 03:17:10 mx2 /netbsd: ahc0: WARNING no command for scb 58 (cmdcmplt)
Feb 12 03:17:10 mx2 /netbsd: QOUTPOS = 239
Feb 12 03:17:10 mx2 /netbsd: sd0(ahc0:0:0:0): SCB 3a - timed out while idle, SEQADDR == 0xa
Feb 12 03:17:10 mx2 /netbsd: SCSIRATE == 0x0
Feb 12 03:17:10 mx2 /netbsd: sd0(ahc0:0:0:0): no longer in timeout, status = 0
Feb 12 03:17:10 mx2 /netbsd: sd0: async, 8-bit transfers, tagged queueing
Feb 12 03:17:10 mx2 /netbsd: sd1: async, 8-bit transfers, tagged queueing
Feb 12 03:17:10 mx2 /netbsd: ahc0: Issued Channel A Bus Reset. 9 SCBs aborted
Feb 12 03:17:10 mx2 /netbsd: sd1: sync (25.0ns offset 63), 16-bit (80.000MB/s) transfers, tagged queueing
Feb 12 03:17:10 mx2 /netbsd: ahc0:A:2: ahc_intr - referenced scb not valid during seqint 0x71 scb(58)
Feb 12 03:17:10 mx2 /netbsd: ahc0: WARNING no command for scb 58 (cmdcmplt)
Feb 12 03:17:10 mx2 /netbsd: QOUTPOS = 0
Feb 12 03:17:10 mx2 /netbsd: sd0: sync (25.0ns offset 63), 16-bit (80.000MB/s) transfers, tagged queueing
Feb 12 03:17:10 mx2 /netbsd: raid3: IO Error.  Marking /dev/sd1f as failed.
Feb 12 03:17:10 mx2 /netbsd: raid3: node (Wsd) returned fail, rolling forward
Feb 12 03:17:10 mx2 /netbsd: raid2: IO Error.  Marking /dev/sd1e as failed.
Feb 12 03:17:10 mx2 /netbsd: raid2: node (Wsd) returned fail, rolling forward
Feb 12 03:17:10 mx2 /netbsd: raid3: node (Wsd) returned fail, rolling forward
Feb 12 03:17:10 mx2 /netbsd: raid0: IO Error.  Marking /dev/sd1a as failed.
Feb 12 03:17:10 mx2 /netbsd: raid0: node (Wsd) returned fail, rolling forward
Feb 12 03:17:10 mx2 /netbsd: raid0: node (Wsd) returned fail, rolling forward
Feb 12 03:17:10 mx2 /netbsd: raid3: node (Wsd) returned fail, rolling forward
Feb 12 03:17:10 mx2 /netbsd: raid3: node (Wsd) returned fail, rolling forward
Feb 12 03:17:48 mx2 /netbsd: sd1(ahc0:0:2:0): SCB 3a - timed out while idle, SEQADDR == 0x9
Feb 12 03:17:51 mx2 /netbsd: SCSIRATE == 0x0
Feb 12 03:17:51 mx2 /netbsd: sd1(ahc0:0:2:0): Queuing a BDR SCB
Feb 12 03:17:51 mx2 /netbsd: ahc0: WARNING no command for scb 58 (cmdcmplt)
Feb 12 03:17:51 mx2 /netbsd: QOUTPOS = 122
Feb 12 03:17:51 mx2 /netbsd: sd1(ahc0:0:2:0): SCB 3a - timed out while idle, SEQADDR == 0xc
Feb 12 03:17:51 mx2 /netbsd: SCSIRATE == 0x0
Feb 12 03:17:51 mx2 /netbsd: sd1(ahc0:0:2:0): no longer in timeout, status = 0
Feb 12 03:17:51 mx2 /netbsd: sd0: async, 8-bit transfers, tagged queueing
Feb 12 03:17:51 mx2 /netbsd: sd1: async, 8-bit transfers, tagged queueing
Feb 12 03:17:51 mx2 /netbsd: ahc0: Issued Channel A Bus Reset. 1 SCBs aborted
Feb 12 03:17:51 mx2 /netbsd: sd0: sync (25.0ns offset 63), 16-bit (80.000MB/s) transfers, tagged queueing
Feb 12 03:17:51 mx2 /netbsd: sd1: sync (25.0ns offset 63), 16-bit (80.000MB/s) transfers, tagged queueing
Feb 12 03:17:51 mx2 /netbsd: ahc0:A:2: ahc_intr - referenced scb not valid during seqint 0x71 scb(58)
Feb 12 03:17:51 mx2 /netbsd: ahc0: WARNING no command for scb 58 (cmdcmplt)
Feb 12 03:17:51 mx2 /netbsd: QOUTPOS = 0
Feb 12 03:18:50 mx2 /netbsd: sd1(ahc0:0:2:0): SCB 3a - timed out while idle, SEQADDR == 0xc
Feb 12 03:18:53 mx2 /netbsd: SCSIRATE == 0x0
Feb 12 03:18:53 mx2 /netbsd: sd1(ahc0:0:2:0): Queuing a BDR SCB
Feb 12 03:18:53 mx2 /netbsd: ahc0: WARNING no command for scb 58 (cmdcmplt)
Feb 12 03:18:53 mx2 /netbsd: QOUTPOS = 34
-------------------------------------- 8< --------------------------------------


Both crashes occurred at 03:15. The /etc/daily script seems to trigger 
it. Building the sources didn't.

Should I send-pr this?

Thanks 
Chris

-- 
----------------------------------------------------------------------
Christoph Kaegi                                           kgc@zhwin.ch
----------------------------------------------------------------------