To (kind-of) follow on from the thread about whether SB2000s were picky
about disks....

Having had no end of pain with one of the disks in my SB2000, which was
eventually replaced with a working one, the other one has now failed. The
difference this time is that I had actually installed SVM and mirrored the disk,
so hopefully I haven't lost anything. (See the metastat output below).

I have swapped the boot disk from c1t2d0 to c1t1d0, which worked well enough, by
changing boot-device from the failed disk to the other one in the OBP, and the
system is now back up, booted from the good disk.

Now ... I've never done this before, so how do I replace the failed mirror? I
assume I can't use the metareplace that metastat suggests, because I only have
two disks. I assume it's something along the lines of;

- metadetach the failed mirrors,
- power down,
- swap the failed disk for a new one
- reboot (will I be able to boot with only one disk?),
- partition the new disk,
- create the metadb's on it
- metaattach the new mirrors?

(I'm coming to hate this damn computer, good learning experience or not.)



d36: Mirror
Submirror 0: d26
State: Needs maintenance
Submirror 1: d16
State: Okay
Pass: 1
Read option: roundrobin (default)
Write option: parallel (default)
Size: 62317824 blocks (29 GB)

d26: Submirror of d36
State: Needs maintenance
Invoke: metareplace d36 c1t2d0s6 <new device>
Size: 62317824 blocks (29 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t2d0s6 0 No Maintenance Yes


d16: Submirror of d36
State: Okay
Size: 62317824 blocks (29 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t1d0s6 0 No Okay Yes


d34: Mirror
Submirror 0: d24
State: Needs maintenance
Submirror 1: d14
State: Okay
Pass: 1
Read option: roundrobin (default)
Write option: parallel (default)
Size: 10247232 blocks (4.9 GB)

d24: Submirror of d34
State: Needs maintenance
Invoke: metareplace d34 c1t2d0s4 <new device>
Size: 10247232 blocks (4.9 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t2d0s4 0 No Maintenance Yes


d14: Submirror of d34
State: Okay
Size: 10247232 blocks (4.9 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t1d0s4 0 No Okay Yes


d33: Mirror
Submirror 0: d23
State: Needs maintenance
Submirror 1: d13
State: Okay
Pass: 1
Read option: roundrobin (default)
Write option: parallel (default)
Size: 21512064 blocks (10 GB)

d23: Submirror of d33
State: Needs maintenance
Invoke: metareplace d33 c1t2d0s3 <new device>
Size: 21512064 blocks (10 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t2d0s3 0 No Maintenance Yes


d13: Submirror of d33
State: Okay
Size: 21512064 blocks (10 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t1d0s3 0 No Okay Yes


d31: Mirror
Submirror 0: d21
State: Needs maintenance
Submirror 1: d11
State: Okay
Pass: 1
Read option: roundrobin (default)
Write option: parallel (default)
Size: 8191680 blocks (3.9 GB)

d21: Submirror of d31
State: Needs maintenance
Invoke: metareplace d31 c1t2d0s1 <new device>
Size: 8191680 blocks (3.9 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t2d0s1 0 No Maintenance Yes


d11: Submirror of d31
State: Okay
Size: 8191680 blocks (3.9 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t1d0s1 0 No Okay Yes


d30: Mirror
Submirror 0: d20
State: Needs maintenance
Submirror 1: d10
State: Okay
Pass: 1
Read option: roundrobin (default)
Write option: parallel (default)
Size: 20484288 blocks (9.8 GB)

d20: Submirror of d30
State: Needs maintenance
Invoke: metareplace d30 c1t2d0s0 <new device>
Size: 20484288 blocks (9.8 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t2d0s0 0 No Maintenance Yes


d10: Submirror of d30
State: Okay
Size: 20484288 blocks (9.8 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t1d0s0 0 No Okay Yes


d35: Mirror
Submirror 0: d25
State: Needs maintenance
Submirror 1: d15
State: Okay
Pass: 1
Read option: roundrobin (default)
Write option: parallel (default)
Size: 20484288 blocks (9.8 GB)

d25: Submirror of d35
State: Needs maintenance
Invoke: metareplace d35 c1t2d0s5 <new device>
Size: 20484288 blocks (9.8 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t2d0s5 0 No Maintenance Yes


d15: Submirror of d35
State: Okay
Size: 20484288 blocks (9.8 GB)
Stripe 0:
Device Start Block Dbase State Reloc Hot Spare
c1t1d0s5 0 No Okay Yes


Device Relocation Information:
Device Reloc Device ID
c1t2d0 Yes id1,ssd@n2000002037d062e2
c1t1d0 Yes id1,ssd@n20000000871a2049



--
"Religion poisons everything."
[email me at huge {at} huge (dot) org <dot> uk]

Huge wrote:

> Having had no end of pain with one of the disks in my SB2000,


I hope I dont follow you! I've just bought a Blade 2000 (2 x 1.2 GHz, 2
x 73 GB, 8 GB RAM) which should be arriving next week.

I've not really thought about what to do with the disks - I'll ponder
that while waiting for it to arrive.

dave

Huge wrote:
> To (kind-of) follow on from the thread about whether SB2000s were picky
> about disks....
>
> Having had no end of pain with one of the disks in my SB2000, which was
> eventually replaced with a working one, the other one has now failed. The
> difference this time is that I had actually installed SVM and mirrored the disk,
> so hopefully I haven't lost anything. (See the metastat output below).
>
> I have swapped the boot disk from c1t2d0 to c1t1d0, which worked well enough, by
> changing boot-device from the failed disk to the other one in the OBP, and the
> system is now back up, booted from the good disk.
>
> Now ... I've never done this before, so how do I replace the failed mirror? I
> assume I can't use the metareplace that metastat suggests, because I only have
> two disks. I assume it's something along the lines of;
>
> - metadetach the failed mirrors,
> - power down,
> - swap the failed disk for a new one
> - reboot (will I be able to boot with only one disk?),
> - partition the new disk,
> - create the metadb's on it
> - metaattach the new mirrors?
>

You will need to do:

1. only for non-Sun disks:
format -> label

2.
format -> partition
If the partition table should be a clone from the other disk,
select the other disk first, in the partition menu give it a name.
Then select the new disk, in the partition menu select the name.

3.
create the metadb's on it

4.
metareplace -e mirror device



--
Michael Tosch @ hp : com

Michael Tosch <eedmit@no.eed.spam.ericsson.pls.se> wrote:
> 2.
> format -> partition
> If the partition table should be a clone from the other disk,
> select the other disk first, in the partition menu give it a name.
> Then select the new disk, in the partition menu select the name.

prtvtoc /dev/rdsk/<source disk>s2 | fmthard -s - /dev/rdsk/<target disk>s2


--
Daniel

Daniel Rock wrote:
> Michael Tosch <eedmit@no.eed.spam.ericsson.pls.se> wrote:
>> 2.
>> format -> partition
>> If the partition table should be a clone from the other disk,
>> select the other disk first, in the partition menu give it a name.
>> Then select the new disk, in the partition menu select the name.
>
> prtvtoc /dev/rdsk/<source disk>s2 | fmthard -s - /dev/rdsk/<target disk>s2
>
>
Huge dont forget to delete the metadb's from the failing disk.
metadb ...
metadb -f -d broken_drive
/Jorgen

On 2007-10-04, Dave <sorry-no-email@nowhere.com> wrote:
> Huge wrote:
>
>> Having had no end of pain with one of the disks in my SB2000,
>
>
> I hope I dont follow you! I've just bought a Blade 2000 (2 x 1.2 GHz, 2
> x 73 GB, 8 GB RAM) which should be arriving next week.
>
> I've not really thought about what to do with the disks - I'll ponder
> that while waiting for it to arrive.

Oh, I always mirror mine - this is the first time I've had to recover, though.


--
"Religion poisons everything."
[email me at huge {at} huge (dot) org <dot> uk]

On 2007-10-04, Michael Tosch <eedmit@NO.eed.SPAM.ericsson.PLS.se> wrote:
> Huge wrote:
>> To (kind-of) follow on from the thread about whether SB2000s were picky
>> about disks....

[snippage]

>> two disks. I assume it's something along the lines of;
>>
>> - metadetach the failed mirrors,
>> - power down,
>> - swap the failed disk for a new one
>> - reboot (will I be able to boot with only one disk?),
>> - partition the new disk,
>> - create the metadb's on it
>> - metaattach the new mirrors?
>>
>
> You will need to do:
>
> 1. only for non-Sun disks:
> format -> label
>
> 2.
> format -> partition
> If the partition table should be a clone from the other disk,
> select the other disk first, in the partition menu give it a name.
> Then select the new disk, in the partition menu select the name.

Yeah, I know how to do all this...

prtvtoc -> fmthard & so on.

>
> 3.
> create the metadb's on it
>
> 4.
> metareplace -e mirror device

So it's OK to metareplace a disk with itself (as it were...)?


--
"Religion poisons everything."
[email me at huge {at} huge (dot) org <dot> uk]

On 2007-10-05, Jorgen Moquist <jorgen.moquist@n.o.s.p.a.m.mailbox.swipnet.se> wrote:
> Daniel Rock wrote:
>> Michael Tosch <eedmit@no.eed.spam.ericsson.pls.se> wrote:
>>> 2.
>>> format -> partition
>>> If the partition table should be a clone from the other disk,
>>> select the other disk first, in the partition menu give it a name.
>>> Then select the new disk, in the partition menu select the name.
>>
>> prtvtoc /dev/rdsk/<source disk>s2 | fmthard -s - /dev/rdsk/<target disk>s2
>>
>>
> Huge dont forget to delete the metadb's from the failing disk.
> metadb ...
> metadb -f -d broken_drive

Had to do that to boot up this morning.

--
"Religion poisons everything."
[email me at huge {at} huge (dot) org <dot> uk]

Huge wrote:
> On 2007-10-04, Michael Tosch <eedmit@NO.eed.SPAM.ericsson.PLS.se> wrote:
>> Huge wrote:
>>> To (kind-of) follow on from the thread about whether SB2000s were picky
>>> about disks....
>
> [snippage]
>
>>> two disks. I assume it's something along the lines of;
>>>
>>> - metadetach the failed mirrors,
>>> - power down,
>>> - swap the failed disk for a new one
>>> - reboot (will I be able to boot with only one disk?),
>>> - partition the new disk,
>>> - create the metadb's on it
>>> - metaattach the new mirrors?
>>>
>> You will need to do:
>>
>> 1. only for non-Sun disks:
>> format -> label
>>
>> 2.
>> format -> partition
>> If the partition table should be a clone from the other disk,
>> select the other disk first, in the partition menu give it a name.
>> Then select the new disk, in the partition menu select the name.
>
> Yeah, I know how to do all this...
>
> prtvtoc -> fmthard & so on.
>
>> 3.
>> create the metadb's on it
>>
>> 4.
>> metareplace -e mirror device
>
> So it's OK to metareplace a disk with itself (as it were...)?
>
>

Is this a question?
Quote from
man metareplace
....
-e Transitions the state of component to the
available state and resyncs the failed com-
ponent. If the failed component has been hot
spare replaced, the hot spare is placed in
the available state and made available for
other hot spare replacements. This command
is useful when a component fails due to

SunOS 5.10 Last change: 8 Aug 2003 1

System Administration Commands metareplace(1M)

human error (for example, accidentally turn-
ing off a disk), or because the component
was physically replaced. In this case, the
replacement component must be partitioned to
match the disk being replaced before running
the metareplace command.


--
Michael Tosch @ hp : com

On 2007-10-05, Michael Tosch <eedmit@NO.eed.SPAM.ericsson.PLS.se> wrote:
> Huge wrote:


>> So it's OK to metareplace a disk with itself (as it were...)?
>>
>>
>
> Is this a question?

Well, thank you for the answer, but I'm given to understand that this piece of
punctuation; "?" means that a question is being asked. Obviously, testing ones
understanding of the man pages isn't allowed. Silly me.


--
"Religion poisons everything."
[email me at huge {at} huge (dot) org <dot> uk]