Jeff, All, We have just started using some new machines with The Intel ESB2 chipset in them (PCI Id: 8086:2681) for the SATA Controller. With any kernels above 2.6.15 (as far as I can tell from my testing at least) we are getting problems with them. This is the output I have managed to collect from the server (Its hard to get much more since the system gets more and more unresponsive as these errors progress): <3>ata2.00: exception Emask 0x10 SAct 0x0 SErr 0x80002 action 0x2 frozen <3>ata2.00: (irq_stat 0x08000000, interface fatal error) <3>ata2.00: tag 0 cmd 0x35 Emask 0x10 stat 0x50 err 0x0 (ATA bus error) <6>ata2: soft resetting port <3>ata2: softreset failed (1st FIS failed) <4>ata2: softreset failed, retrying in 5 secs <6>ata2: hard resetting port <4>ata2: port is slow to respond, please be patient I have found some other people with similar problems and this is their output: ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen ata2.00: tag 0 cmd 0xb0 Emask 0x4 stat 0x40 err 0x0 (timeout) ata2: soft resetting port ata2: softreset failed (port busy but CLO unavailable) ata2: softreset failed, retrying in 5 secs ata2: hard resetting port ata2: port is slow to respond, please be patient (Status 0x80) ata2: port failed to respond (30 secs, Status 0x80) ata2: COMRESET failed (device not ready) ata2: hardreset failed, retrying in 5 secs ata2: hard resetting port ata2: SATA link up 1.5 Gbps (SStatus 113 SControl 300) ata2.00: configured for UDMA/133 ata2: EH complete SCSI device sdb: 781422768 512-byte hdwr sectors (400088 MB) sdb: Write Protect is off sdb: Mode Sense: 00 3a 00 00 SCSI device sdb: drive cache: write back This second bunch of output matches mine pretty much word for word (except different size disks) as far as I can tell! I am trying to get a detailed output from the machine from the console server and will follow up with this and forward on my exact logs if I can get some. Interestingly I only get this behaviour when I am doing software RAID on the machine (not via the hardware but rather the Linux MD support). I don't know if this is useful information, I don't see the problem when running non-mirrored disks or when just doing anything like the following: dd if=/dev/urandom of=/dev/sd[a|b] bs=1024k. Even to both disks at the same time. Can you throw any light on this at all? Suggest any work around? I'm happy to try and provide any other information I can. -- James Ray. <j.ray@xxxxxxxxxx> Computing Services Queen Mary, University of London - 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
