RAID is a matter of availability, not data security
Hubert Tonneau (hubert.tonneau@heliosam.fr)
Tue, 24 Aug 1999 17:54:20 +0200
Having a perfeclty stable kernel that requires to umount the file system (so
stop the real service) each time a hardware failure appends is just
hypocritical: users don't care at all if the problem is OS related or
hardware related; they care of availability.
RAID IS A MATTER OF PROVIDING MORE AVAILABILITY FROM STANDARD HARDWARE,
and it requires hot repair, so if you want to keep the old code in the kernel,
call it toy raid, or call the new one real raid, but put the 0.90 code in the
2.2 standard kernel, eventualy with EXPERIMENTAL in front of it,
or never pretend that 2.2 is a stable kernel: it's just an OS designers
masturbation.
Or would it mean that enabling SUPPORT_RECONSTRUCTION in the old code is
considered safer than the new code and is maintained more actively ?
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/