Ishikawa wrote:
>
> Hi,
>
> Be it per-cotroller, or per-device, I think it is great
> to have the timeout variable controllable (possibly) from
> userland programs (via ioctls and such.)
>
> I believe that Alan Cox bumped the timeout to 30 seconds
> after hearing complaint about too short timeout
> from people using RAID (and I also chimed in saying
> that old disks tend to take a long time when read
> error occurs and recovery is attempted.).
It was bumped to 30 because the Megaraid controllers were getting resets when
there was no need for them. This is a double edged sword. I've had many
people since then complain to me about how long it takes for a reset to
complete when something does go wrong on the aic7xxx driver (I know, it
shouldn't happen, but tough crap, the world changes, the aic7xxx based
controllers change, and the devices change, aka, shit happens). Making it per
device selectable means that devices that we know should timeout quickly, like
regular hard disks, can have a low timeout number while the RAID devices can
have a high timeout number.
> I think this is in the right direction, and
> many other timeout values ought to be variables, I think.
>
> Happy Hacking,
>
> Chiaki Ishikawa
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> the body of a message to [EMAIL PROTECTED]
--
Doug Ledford <[EMAIL PROTECTED]> http://people.redhat.com/dledford
Please check my web site for aic7xxx updates/answers before
e-mailing me about problems
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.tux.org/lkml/
