On Sat, 12 Jan 2002, Greg Kettmann wrote: > The fstab entry = LABEL=/usr /usr ext2 defaults 1 2. I don't quite get > that "LABEL" thing.
ext2fs filesystems (and others) can have a filesystem "label", i.e., an arbitrary name. By putting "LABEL=whatever" in /etc/fstab, mount(8) knows to scan partitions for an ext2fs filesystem with label "whatever", and mount that. Among other things, this makes the system less susceptible to drive IDs changing. Apparently, the filesystem you mount on /usr is labeled "/usr". > After installing this drive I ran a read check, the only extensive check > I can run, from within the Adaptec menu's. Both drives run clean. I have seen bad drives pass that check before, so I am of the opinion that it is only good for telling you something is wrong, not confirming everything is right. > SCSI host 0 abort (pid 0) timed out - resetting > SCSI bus is being reset for host 0 channel 0 > I/O error: dev 08:01, sector 16 > I/O error: dev 08:01, sector 80 Like it says, the command timed out. The driver sent a command to the host adapter, and the command never completed. This is bad news, and usually indicates a hardware problem. However, it can also be a bad device driver -- when I tried an early 2.4 kernel on my system at home (with an AHA-2940UW), I got all sorts of bogus SCSI errors. 2.2 runs fine. > Any thoughts or suggestions here. I've changed the devices postition on > the ribbon cable, although I always keep on device as the last one on > the cable. Use an external, active terminator. Not only does it result in more reliable operation, it makes changing things around easier, since you don't have to futz with termination on the devices. If you are using a 68-to-50-pin cable adapter, make sure it terminates the wide pins. (Actually, if possible, get rid of it -- while it is possible to make those things correctly, there are an awful lot of cheap SCSI components out there.) (This does not apply if you only have 50-pin cables everywhere, and are simply attaching 68-pin devices to it.) > Since I can boot the system (from the ide drive) I can try bringing up > one device at a time. However, does someone have a suggestion for a way > to stress a drive, say a big copy or a write and read sequence? badblocks(8) > It really, really looks like a hardware problem however since the > adaptec read test works I'm wondering if it could be a Linux Driver > problem. Adaptec's site doesn't help. What other resources might I > find helpful. Kernel source. The Adaptec drivers have READMEs, and the source code comments can be useful. You can also try different drivers. In 2.4, there is the "classic" aic78xx driver, written by the kernel folks, and the "new" aic78xx driver, written by Adaptec. The 2.2 driver is different from both of those. One may work better than the others. Google. ;-) -- Ben Scott <[EMAIL PROTECTED]> | The opinions expressed in this message are those of the author and do not | | necessarily represent the views or policy of any other person, entity or | | organization. All information is provided without warranty of any kind. | ***************************************************************** To unsubscribe from this list, send mail to [EMAIL PROTECTED] with the text 'unsubscribe gnhlug' in the message body. *****************************************************************
