Heinz Mauelshagen wrote:
On Sat, Nov 27, 2004 at 12:33:24AM +0100, Matthijs Melchior wrote:
L.S.,
The RAID config on my Dell 8400, as set by BIOS and Win-XP is: --------------------------------------------------------------
Intel(R) Application Accelerator RAID Option ROM v4.0.0.6211 Copyright(C) 2003-04 Intel Corporation. All Rights Reserved.
RAID Volumes:
ID Name Level Strip Size Status Bootable
0 ARRAY RAID1(Mirror) N/A 149.0GB Normal Yes
1 ARRAY2 RAID1(Mirror) N/A 74.5GB Degraded Yes
2 ARRAY3 RAID3(Stripe) 128KB 149.0GB Failed No
Physical Disks:
Port Drive Model Serial # Size Type/Status(Vol ID)
0 WDC WD1600JD-75H WD-WMAL91264840 149.0GB Member Disk(0)
1 WDC WD1600JD-00H WD-WMAL91985532 149.0GB Member Disk(1,2)
2 WDC WD1600JD-75H WD-WMAL91409154 149.0GB Member Disk(0)
3 WDC WD1600JD-00H WD-WMAL91887644 149.0GB Non-RAID Disk
Press <CTRL-I> to enter Configuration Utility.........
------------------------------------------------------------------------------
RAID volume 0 is a RAID1 set of two full disks, port 0 and port 2
RAID volume 1 is a RAID1 set of half of two disks, port 1 and port 3
RAID volume 2 is a RAID0 set of the orher half of two disks, port 1 and port 3
The RAID info on disk 3 has been erased, which causes the Degraded and Failed
states for volumes 1 and 2. Linux has been installed on disk 3.
The Linux system is running kernel 2.6.10-rc1, which has a working ahci module.
'dmraid' has the following view on this machine:
++ dmraid -V dmraid version: 1.0.0-rc5f (2004.11.24) dmraid library version: 1.0.0-rc5f (2004.11.24) device-mapper version: 4.1.0 ++ dmraid -b /dev/sda: 312500000 total, "WD-WMAL91264840" /dev/sdb: 312581808 total, "WD-WMAL91905532" /dev/sdc: 312500000 total, "WD-WMAL91409154" /dev/sdd: 312581808 total, "WD-WMAL91887644" ++ dmraid -r /dev/sda: isw, "isw_ebdcfejfgd", GROUP, ok, 312499998 sectors, data@ 0 /dev/sdb: isw, "isw_dbiacddhid", GROUP, ok, 312581805 sectors, data@ 0 /dev/sdc: isw, "isw_ebdcfejfgd", GROUP, ok, 312499998 sectors, data@ 0 ++ dmraid -s -g isw: unsupported map state 0x2 on /dev/sdb for ARRAY2 ERROR: adding /dev/sdb to RAID set "isw_dbiacddhid" ERROR: removing RAID set "isw_dbiacddhid" *** Superset name : isw_ebdcfejfgd size : 624999996 stride : 0 type : GROUP status : ok subsets: 1 devs : 2 spares : 0 --> Subset name : isw_ebdcfejfgd_ARRAY size : 312499200 stride : 256 type : mirror status : ok subsets: 0 devs : 2 spares : 0
------------------------------------------------------------------------------
It is unfortunate dmraid does not yet know about degraded RAID1 sets....
I want to help getting support for this, and for resyncing the device, in dmraid.
Please give me some hints on how to proceed.
Matthijs,
see the metadata definitions for raid_dev and raid_set, how those get
silled in in isw.c and used by activate.c.
OK, I will investigate that....
Support for degraded RAID1 set resynchronization can be worked in, support for AAID3 not because of no RAID3 device-mapper target yet.
It would be a start to just activate a degraded RAID1 set with just 1 disk, and not remove it.
I do not understand where this remark about RAID3 is comming from, currently, I am only interrested in RAID0 and RAID1......
Heinz
..........
-- Regards, ---------------------------------------------------------------- -o) Matthijs Melchior Maarssen /\\ mmelchior@xxxxxxxxx Netherlands _\_v ---------------------------------------------------------------- ----
