On 7.6.2000 Alan Cox wrote:
> > If you try doing that with ATA you would need 7 or 8 free slots
> > _just_for_drive_controllers_. Not to mention interrupts and interrupt
> > loading (you would need a much faster CPU to deal with the IDE case).
>
> Actually you need one slot per 16 drives with an 8 port 3ware controller
Make that one slot per 8 drives currently: 3ware wants all
drives configured as master, 1 drive per cable.
BTW, be careful if you consider using the 3ware boards in raid1(mirror)
mode,
there's a bug in the current firmware: It does not catch unclean shutdown
in case of power loss, reset, kernel crash or whatever. If a write is in
progress during such an event, you'll get inconsistent data on the disks -
one will have new data already written, the other write may not have
finished.
This is why software raid will do a complete resync of data on next startup
in
this case.
However, on next reboot the controller does _NOT_ recognice this
inconsistency. It happily uses both disks as a healty mirror. If you now do
an md5 sum of the file you were last writing to and repeatedly check the
md5sum, you'll get inconsistant results - 50% of the time, the file will
check OK and 50% there'll be a checksum error since the reads are
distributed over both disks to increase performance.
Bye, Martin
"you have moved your mouse, please reboot to make this change take effect"
--------------------------------------------------
Martin Bene vox: +43-316-813824
simon media fax: +43-316-813824-6
Andreas-Hofer-Platz 9 e-mail: mb@sime.com
8010 Graz, Austria
--------------------------------------------------
finger mb@mail.sime.com for PGP public key
-
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/
This archive was generated by hypermail 2b29 : Wed Jun 07 2000 - 21:00:28 EST