Disk controller reassignment


 
Thread Tools Search this Thread
Operating Systems Solaris Disk controller reassignment
# 8  
Old 06-08-2013
Speaking from the experience with Netra T4-2: when you configure hardware RAID, the resulting disk in "format" contains "LSI" in its name, which is not the case here. Also, when two disks are hardware-mirrored, they are visible as single disk in Solaris, so there would be 7 disks visible here and not 8.
# 9  
Old 06-08-2013
@bartus11....I agree with you but what possible explanation is there for Solaris to decide to mirror a 300GB drive to a 600GB drive, all on its own on a new build???

How does that possibly work???

The only way that I can think of is a pre-existing hardware RAID configuration.
# 10  
Old 06-08-2013
To understand the situation, we should first get the output of
Code:
zpool status
metastat -a

# 11  
Old 06-08-2013
Is raidctl works on that platform?
If yes, then it could be helpful too.
# 12  
Old 06-09-2013
@GP81.....

I don't think 'raidctl' will work on a T4-1. Raidctl is for the older built-in LSI hardware raid controllers used on hardware like the T2000.

If it was a LSI like that, then as Bartus11 says, once a raid mirror is configured the mirror drive is 'hidden' from view.

This is a T4-1 which uses a different raid controller and different command set.......

Create a Hardware RAID Volume - SPARC T4-1 Server HTML Document Collection

I've already made the point that typical factory configurations have the first two slots mirrored when supplied to the customer and this appears to be what we've got here.

We've got to the point where several members are trying to help and speculating over the resolution of the problem but what we need is Kerrygold to post more output.

I agree with Bartus11 that we all wait for Kerrygold to get back online otherwise we are all just guessing.

My own experience with T4 series is pretty thin anyway.
# 13  
Old 06-10-2013
Hi,

Thanks for the help so far.
The problem is seen just after the solaris is installed (it tries to create the root mirror on the 600gb disk instead of the 300gb)

the output of metastat is:
Code:
# metastat d0
d0: Mirror
    Submirror 0: d10
      State: Okay         
    Submirror 1: d20
      State: Resyncing    
    Resync in progress: 0 % done
    Pass: 1
    Read option: roundrobin (default)
    Write option: parallel (default)
    Size: 542925000 blocks (258 GB)

d10: Submirror of d0
    State: Okay         
    Hot spare pool: hsp001
    Size: 542925000 blocks (258 GB)
    Stripe 0:
        Device                             Start Block  Dbase        State Reloc Hot Spare
        /dev/dsk/c0t5000CCA03C3B0B70d0s0          0     No            Okay   Yes 


d20: Submirror of d0
    State: Resyncing    
    Hot spare pool: hsp001
    Size: 1048576968 blocks (500 GB)
    Stripe 0:
        Device                             Start Block  Dbase        State Reloc Hot Spare
        /dev/dsk/c0t5000CCA03C3F98F8d0s0          0     No            Okay   Yes 


Device Relocation Information:
Device                           Reloc  Device ID
/dev/dsk/c0t5000CCA03C3B0B70d0   Yes    id1,sd@n5000cca03c3b0b70
/dev/dsk/c0t5000CCA03C3F98F8d0   Yes    id1,sd@n5000cca03c3f98f8
#

I will try to break mirrors manually with raid tools and see what i can do...
# 14  
Old 06-10-2013
Personally I've never seen a Solaris install routine set up a mirror all on its own. Okay, I guess you learn something new every day!!!

However, the metastat output confirms that is what it's done so, as you say, you can break the mirror and remake the mirror with a different drive using the tools. Give it a go.
Login or Register to Ask a Question

Previous Thread | Next Thread

10 More Discussions You Might Find Interesting

1. Solaris

Sun Fire v440 Hard disk or controller broken? WARNING: /pci@1f,700000/scsi@2/sd@0,0 (sd1)

Hi, I have a Sun Fire V440 server that fails to boot up correctly. A lot of services are not started and the sytems acts really slow to commands. During boot I can see the following Error: WARNING: /pci@1f,700000/scsi@2/sd@0,0 (sd1): SCSI transport failed: reason 'reset': retrying... (15 Replies)
Discussion started by: oliwei
15 Replies

2. Solaris

Changed disk controller -> drops into GRUB?!

I changed disk controller. Unplugged every disk, inserted another controller card, an IBM M1015. The disks are recognized in BIOS, and the controller card's BIOS shows up. But after that, I drop into GRUB and a prompt. That is all. How shall I solve this problem? I am thinking of booting from... (2 Replies)
Discussion started by: kebabbert
2 Replies

3. SCO

Change Disk controller in SCO 6.0.0

So today I tried to upgrade a customer from a Proliant with an HPSAS controller to an Intel server with a MEGASAS controller. First, I downloaded the megasas driver from SCO and installed it on the old system. Then I created a new Microlite Edge emergency boot cd with the new driver on it. I... (2 Replies)
Discussion started by: jgt
2 Replies

4. Solaris

Help needed to find out the disk controller for veritas disks

hi all i am using vxvm 5.1 on my sun blade 150 which is running with solaris 5.10. When i give the command "vxdisk list" it gives the following output # vxdisk list DEVICE TYPE DISK GROUP STATUS c0t0d0s2 auto:none - - online... (2 Replies)
Discussion started by: kingston
2 Replies

5. SCO

SCO 5.0.7 no root disk controller found error during install

I'm "attempting" to install SCO 5.0.7 on an HP ML370 G4 server and am ready to bash the keyboard with head now. I keep getting the error message "WARNING hd: no root disk controller found" when running the bootable install cd. I have a raid 5 array with an online spare created using 4 36.3 GB... (2 Replies)
Discussion started by: FrictionBurn
2 Replies

6. HP-UX

[Joy Share] - Replace an unknown status controller on HP FC60 disk array

Replace an unknown status controller on HP FC60 disk array Author:JoyOnline JoyChangLu@gmail.com Trouble description: The users detected that the status of controller at A is unknown. Analysis and diagnostic: Physical observed all leds on enclosure of disk array is normal. The... (4 Replies)
Discussion started by: JoyOnLine
4 Replies

7. SCO

Warning HD: no root disk controller was found

hi... i hav a problem to install sco unix while installing sco unix 5.0.5 in acer altos with hardware raid am getting error "Warning HD: no root disk controller was found system error 5: while getting fs type flushing controller: DAC960PRL cache done." i hav use the following... (18 Replies)
Discussion started by: mselvaprakash
18 Replies

8. Solaris

changing controller name

Dear Solarizer, If i have a disk let say like this c1t0d0s0 How to change the controller to be come example : c3t0d0s0 Thanxs For your replay Regads, tpx (3 Replies)
Discussion started by: tpx99
3 Replies

9. Solaris

Controller target disk?

Hi, I am a newbie to Solaris. I want to know how to find out the controller target and disk number on a SunFire V890 box that has 6X72GB disks. The probe-scsi output is as shown below: /pci@8,600000/SUNW,qlc@2 LiD HA LUN --- Port WWN --- ----- Disk description ----- 0 0 0 ... (1 Reply)
Discussion started by: sudhir_shet
1 Replies

10. UNIX for Advanced & Expert Users

Controller Naming

Hello all, How does the Solaris identifies the controller subscript ? ( like c0txdxs0 or c1txdxsx ?? ) I have a unix box ( Ultra 30) running with 2.5.1. When I connected an external hard disk to the on-board scsi port, it got identified as c0t1dxsx... (... (1 Reply)
Discussion started by: shibz
1 Replies
Login or Register to Ask a Question