Tejun Heo writes: > Mike Accetta wrote: > > The version of hdparm on this box gives > > dev/sdb: > > operation not supported on SCSI disks > > against the SCSI disk name. With the controller in compatible mode, > > here is the 'hdparm -I /dev/hdb' output on the same disk: > > > > ATA device, with non-removable media > > Model Number: FUJITSU MHT2060BH > > Serial Number: NR0YT53258G8 > > Firmware Revision: 0000104A > > Sorry about late late late response. :-) > > 2.6.21 will behave better and turn off NCQ automatically in such cases. > It also reports proper model and firmware revision, so we'll be able to > blacklist it from info contained in the dmesg. > > Does your drive always work reliably if queue depth is set to one? And > often fails when queue depth is 31? Hey, better late than never :-) The drive actually seems to work fine at queue depth 4 and that is what we are currently running it at. I stopped at 4 when that worked and never went any lower. It always fails above 4 eventually. The larger the number, the quicker it seems to fail. A raid re-resync operation was the trigger. On a 60Gb disk, I don't think I ever saw it complete without error at NCQ > 4. Sometimes the re-sync would logically succeed, but with timeout errors, retries and downshifts logged to the console. Other times, it would get the timeouts and do the retry/downshift until it ran out of modes and then seem to get stuck in a retry loop. -- Mike Accetta ECI Telecom Ltd. Data Networking Division (previously Laurel Networks) - To unsubscribe from this list: send the line "unsubscribe linux-ide" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
