Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 4 Apr 2003 18:21:50 -0500
From:      "Michael G. Jung" <mikej@confluenttech.com>
To:        <freebsd-fs@freebsd.org>
Subject:   5.0 Raidframe / SMP / Questions
Message-ID:  <9D7F0DF3FB16D41184010050DA90E000BF6631@neo.confluentasp.local>

next in thread | raw e-mail | index | archive | help
First I'm new to raidframe... but I have a few questions/problems - =
thanks to all
in advance...

--mikej
Michael Jung

The scenario - I have 20xFastWide 18.2 disks - would like to mirror two =
for boot=20
2 hot spare and use 16 raid 5(4?).... I'm looking for drive space and =
not necessarily=20
raw speed.

I'm testing with:=20

FreeBSD dual-450.mikej.com 5.0-RELEASE FreeBSD 5.0-RELEASE #0: Thu Jan =
16 22:16:53 GMT 2003     =
root@hollin.btc.adaptec.com:/usr/obj/usr/src/sys/GENERIC  i386

0) I've been reading the netbsd list and it seems problems exist with =
smp... is that=20
   the case with FreeBSD ????

1) "man raidcontrol" indicates that you need to set the disklabel to =
"RAID" if you want
    the kernel to boot from the partition but disklabel stubbornly =
insists that "RAID" is not a valid
    type...  I get this error when trying to do so with disklabel......

   "line 23: Warning, unknown file system type RAID"

   and it kindly marks the partion type to "unknown"

2) Does RAID 4 work?????  for write performance would this not be a =
better option?

3)  I have two drives in the chain that I get "unknown IOCTL" when they =
are in my raid
    config... yet they are labeled exactly like the rest.....  this is =
when running raidctl -C myconfigfile
   (yeah  da3s1e and da9s1e are the culprits)


Here is the config file (note they are commented out.. and the drive =
count is 8 active drives)

START array
# numRow numCol numSpare
1 8 0

START disks
/dev/da0s1e
/dev/da1s1e
/dev/da2s1e
#/dev/da3s1e
/dev/da4s1e
/dev/da5s1e
/dev/da6s1e
/dev/da7s1e
/dev/da8s1e

START spare
#/dev/da9s1e

START layout
# sectPerSU SUsPerParityUnit SUsPerReconUnit RAID_level_5
32 1 1 5

START queue
fifo 100

++++++++++++++++++++
Here is a sample disklabel of a drive that complains about IOCTL
++++++++++++++++++++

# /dev/da3s1c:
type: SCSI
disk: da3s1
label:=20
flags:
bytes/sector: 512
sectors/track: 63
tracks/cylinder: 255
sectors/cylinder: 16065
cylinders: 2213
sectors/unit: 35566480
rpm: 3600
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0           # milliseconds
track-to-track seek: 0  # milliseconds
drivedata: 0=20

8 partitions:
#        size   offset    fstype   [fsize bsize bps/cpg]
  c: 35566480        0    unused        0     0         # (Cyl.    0 - =
2213*)
  e: 35551782        0    4.2BSD        0     0     0   # (Cyl.    0 - =
2212*)
(root@dual-450) /etc#=20


++++++++++++++++++++++
Here is a drive that works in my config file
++++++++++++++++++++++

(root@dual-450) /etc# disklabel da1s1
# /dev/da1s1c:
type: SCSI
disk: da1s1
label:=20
flags:
bytes/sector: 512
sectors/track: 63
tracks/cylinder: 255
sectors/cylinder: 16065
cylinders: 2213
sectors/unit: 35566480
rpm: 3600
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0           # milliseconds
track-to-track seek: 0  # milliseconds
drivedata: 0=20

8 partitions:
#        size   offset    fstype   [fsize bsize bps/cpg]
  c: 35566480        0    unused        0     0         # (Cyl.    0 - =
2213*)
  e: 35551782        0    4.2BSD        0     0     0   # (Cyl.    0 - =
2212*)
(root@dual-450) /etc#=20


NOTE:  da0 has the lowest sector count of 35551782... so I've set all e: =
slices to this value




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9D7F0DF3FB16D41184010050DA90E000BF6631>