Hello,

I've got a DELL PowerEdge 2600 with a dual RAID hardware controller DELL
PERC 4/Di.

Thanks to patches I found in NetBSD-current mailing list, I was able to
use the PERC 4/Di with the amr* driver; cf. :

http://mail-index.netbsd.org/current-users/2003/06/09/0071.html

It was a nightmare to install, especially because of the large size of
the RAID-logical drives ld*, that implied some modification of the usual
scheme for disklabel (block size, fragment size).

% dmesg |grep ld0
ld0 at amr0 unit 0: RAID 1, optimal
ld0: 69878 MB, 8908 cyl, 255 head, 63 sec, 512 bytes/sect x 143110144 sectors
boot device: ld0
root on ld0a dumps on ld0b

% disklabel ld0
# /dev/rld0d:
type: unknown
disk: unknown
label: fictitious
flags:
bytes/sector: 512
sectors/track: 63
tracks/cylinder: 255
sectors/cylinder: 16065
cylinders: 8908
total sectors: 143110144
rpm: 7200
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0           # microseconds
track-to-track seek: 0  # microseconds
drivedata: 0

4 partitions:
#        size    offset     fstype  [fsize bsize cpg/sgs]
   a:   6297417        63     4.2BSD   4096 16384    32   # (Cyl.    0*- 391)
   b:   2088450   6297480       swap                      # (Cyl.  392 - 521)
   c: 143110081        63     unused      0     0         # (Cyl.    0*- 8908*)
   d: 143110144         0     unused      0     0         # (Cyl.    0 - 8908*)


I set empiricaly fsize, bsize and cpg/sgs because newfs was complaining about these values
(and thanks to its output).

Where I can find doc. about how to set these values correctly ? I saw nothing in the NetBSD official documentation :(


I did not found documentation about how to set fsize, bsize and cpg/sgs in order
to be able to use newfs

kernel is 1.6U (2003/06/19) using your patches and system is 1.6.1 ...

some utilies kernel-related do not work of course (lsof, systat), but
I'm a little afraid of running -current on a production server, since I
was not able to build a complete release of -current-2003-07-19.

I was wondering if it's possible to get infos from the Perc 4/Di, such
as status of RAID logical devices : I removed a harddrive of a RAID-5
logical drive, but not kernel msg happened.

the scsictl utility does not work of course, since no scsibus are
attached to the amr* driver.

Do you have some hints about it ?

Regards,

Pierre Bourgin

Tracy Di Marco White a écrit:
 > Pierre Bourgin <pierre.bourgin@pcotech.fr> wrote:
 >
 >>Thanks to your post containing some wonderful patches, I think I will be
 >>able too to use NetBSD on a DELL PowerEdge 2600 (I'm building the
 >>resulting kernel).
 >>
 >>Really thanks for it !!!!
 >
 >
 > You're welcome.
 >
 >
 >>your post:
 >>
 >>http://mail-index.netbsd.org/current-users/2003/06/09/0071.html
 >>
 >>
 >>But do you know why theses patches have not been merged into -current ?
 >>It would be really nice :-)
 >
 >
 > Andrew Doran wrote them, and was working on additional changes to
 > fix what causes the messages you get below.  I'll ask if there's
 > any reason for the delay.
 >
 >
 >>Do you want that I send some additionnal infos. about it ? for instane,
 >>I've got kernel msgs like 'amr0: enquiry failed (st=2)'.
 >
 >
 > I have that disabled, I'm not sure I have the patch around that disables
 > it though.
 >
 > -Tracy
 >
 >