Hello all, I am attempting to install OpenSolaris 2009.06 on a new server and I'm running into the oddest problem.
The server has two LSI SAS cards (SAS3081E) and 16 total drives spread evenly across the two controllers: 0 - 64GB SSD (L2ARC) 1 - 64GB SSD (L2ARC) 2 - 32GB SSD (SLOG) 3 - 1TB HD (64GB rpool, 864GB data zpool) 4 though 7 - 1TB HD (928GB data zpool) Each drive is mirrored with the corresponding drive on the other controller, except the L2ARC drives. So we've got a 4-SSD L2ARC stripe, a 2-SSD SLOG mirror, and a 5-stripe set of mirrored data partitions. No matter what I do, drive 3 (the rpool/zpool drive) on the first controller disappears under load. The following appears in /var/adm/messages when this happens... Once at the top: scsi: [ID 107833 kern.warning] WARNING: /p...@7a,0/pci8086,3...@5/pci1000,3...@0/s...@f,0 (sd26): SYNCHRONIZE CACHE command failed (5) Followed by this repeated 4 times: scsi: [ID 365881 kern.info] /p...@7a,0/pci8086,3...@5/pci1000,3...@0 (mpt0): Log info 0x31080000 received for target 15 scsi_status=0x0, ioc_status=0x804b, scsi_state=0x0 Followed by this repeated many times: scsi: [ID 107833 kern.warning] WARNING: /p...@7a,0/pci8086,3...@5/pci1000,3...@0/s...@f,0 (sd26): Command failed to complete...Device is gone I have swapped drives, swapped controllers, swapped cables, swapped drive carriers and swapped drive bays, and the problem always stays in place (i.e. it is always the split rpool/data drive on the first controller, even if I completely change which physical equipment that is. Does anyone have any ideas as to why this could be happening? I'm absolutely at my wits' end here! :-( Thanks for any suggestions! -- This message posted from opensolaris.org _______________________________________________ opensolaris-help mailing list opensolaris-help@opensolaris.org
