Re: ATA mkIII first official patches - please test!

From: Frode Nordahl <frode_at_nordahl.net>
Date: Sat, 5 Feb 2005 12:16:59 +0100
On Feb 3, 2005, at 21:52, Søren Schmidt wrote:

> ATA-mkIII first official snapshot.

Yes! Just in time! Thanks, Søren :-)

Using kernel from RELENG_5, supped yesterday.

Tried to install on in a box with a Promise TX 2200 card, the bootup 
took forever, but the install worked fine when it finished.

Maby related to the third "phantom" PATA bus on that board? :-)

I'm installing from network (PXE) and for some reason I don't get any 
kernel messages on the console when doing that, so I don't know when or 
where it 'hangs'.

I forgot to copy in my modified kernel after the install, so I never 
got it booting after that, but I'll redo as soon as possible and get 
back with more details about this.

> o       ATA RAID support has been rewritten and and now supports these
>         metadata formats:
>                 "Adaptec HostRAID"
>                 "Highpoint V2 RocketRAID"
>                 "Highpoint V3 RocketRAID"
>                 "Intel MatrixRAID"
>                 "Integrated Technology Express"
>                 "LSILogic V2 MegaRAID"
>                 "LSILogic V3 MegaRAID"
>                 "Promise FastTrak"
>                 "Silicon Image Medley"

This works with my Intel S875WP1-E with Intel ICH5 / Adaptec HostRAID 
onboard!

(The product specs says Promise controller, but that is only when you 
buy the 4-ch SATA version)

However, if I create a RAID1 array in "Quick Mode", installation failes 
with write errors ("disk full"). The Adaptec management BIOS warns 
about creating arrays in Quick Mode, so this may be a problem of 
theirs.

Anyway, is there any way to detect that an array is created in this 
way, and handle it, or warn about it?

The warning message from the BIOS says "You may need to repair the 
array once OS is installed" or similar, so the array probably should be 
marked DEGRADED, and be rebuilded once the installation is finished.


I can no longer use my swap partition as dump device:

# dumpon /dev/ar0s1b
dumpon: ioctl(DIOCSKERNELDUMP): Operation not supported by device

But I'm not certain that it has ever been allowed? :-)

Maybe it won't work to have dump device on RAID, since the kernel won't 
be able to keep the RAID in shape when it's dumping core?


I have not done any extensive testing yet. I'll get these boxes 
installed and set up various tests to see how it performs. I'll also do 
some hotswap testing on the TX2200.


> Enjoy!

I most certainly will! Thanks.


Regards,
Frode Nordahl

> -- 
>
> -Søren
>
>
> _______________________________________________
> freebsd-current_at_freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to 
> "freebsd-current-unsubscribe_at_freebsd.org"
Received on Sat Feb 05 2005 - 10:17:04 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:27 UTC