Spare space problem.

[MammaGutt]

If I do like this:

Code: Select all

setsys SparingAlgorithm Minimal
admithw

setsys SparingAlgorithm Default
admithw

Will it help restore default spare chunks count?

Yes and I would suggest to try that.

If it doesn't fix any spare chunklets, I would do:

removespare a:a
admithw

to simply remove everything and try to recreate.
You should probably wait a minute between the two commands to ensure you've got free chunklets on SSD.

When I look at your number I don't understand anything....

3-par-2 looks correct....
446 chunklets per 480GB SSD, so spare equal 1 drive per 24 drives, minimum 2 drives = 892 chunklets.
1116 chunklets per 1.2TB FC drive, so spare equal 1 drive per 24 drives, minimum 2 drives = 2232
3611 chunklets per 4TB NL drive, so spare equal 1 drive per 24 drives, minimum 2 drives - FC spare chunklets = 7222 - 2232 = 4990.


For 3-par-1 SSD doesn't make any sense no matter how one do the numbers.

[Morbo]

Nothing is helping : (
Nor seting sparealgorithm to minimal and admithw nor tunesys
is it safe to run removespare command?
If I create a map of spare chunks using showspare and manually remove half of these running a lot of commads like this:
removespare 55:193
removespare 55:194
removespare 55:195
....
will it be safe?

[MammaGutt]

Nothing is helping : (
Nor seting sparealgorithm to minimal and admithw nor tunesys
is it safe to run removespare command?
If I create a map of spare chunks using showspare and manually remove half of these running a lot of commads like this:
removespare 55:193
removespare 55:194
removespare 55:195
....
will it be safe?

Everything will be safe, but it will take forever .....

What does admithw say when it passes thru the spare chunklet part of the process?

Is it ignoring everything because it has too many, is it failing and throwing an error message?

[Morbo]

Hmm, i didt notice on first time but seems admithw ends with error

Code: Select all

admithw
Checking nodes...

Checking volumes...

Checking system LDs...

Checking ports...

Checking state of disks...
Checking for drive table upgrade packages

Checking cabling...

Found one or more discrepancies when examining cage connections:

Component --------------Description-------------- Qty
Cabling   Cages not connected to same slot & port   2

-Identifier- -- ------------------------Description------------------------
cage1        -- Cage connected to different ports node0 DP-1 and node1 DP-2
cage2        -- Cage connected to different ports node0 DP-2 and node1 DP-1

Enter c to continue despite this issue or q to quit and fix the issue manually:
c


Checking cage firmware...

Checking if this is an upgrade that added new types of drives...

Checking for disks to admit...
Checking for drive table upgrade packages
0 disks admitted


Checking admin volume...
Admin volume exists.

Checking if logging LDs need to be created...
No new logging LDs need to be created

Checking if preserved data LDs need to be created...
No new preserved data LDs need to be created

Checking if system scheduled tasks need to be created...

Checking if the rights assigned to extended roles need to be updated...
Permission denied

[MammaGutt]

Hmm, i didt notice on first time but seems admithw ends with error

Code: Select all

admithw
Checking nodes...

Checking volumes...

Checking system LDs...

Checking ports...

Checking state of disks...
Checking for drive table upgrade packages

Checking cabling...

Found one or more discrepancies when examining cage connections:

Component --------------Description-------------- Qty
Cabling   Cages not connected to same slot & port   2

-Identifier- -- ------------------------Description------------------------
cage1        -- Cage connected to different ports node0 DP-1 and node1 DP-2
cage2        -- Cage connected to different ports node0 DP-2 and node1 DP-1

Enter c to continue despite this issue or q to quit and fix the issue manually:
c


Checking cage firmware...

Checking if this is an upgrade that added new types of drives...

Checking for disks to admit...
Checking for drive table upgrade packages
0 disks admitted


Checking admin volume...
Admin volume exists.

Checking if logging LDs need to be created...
No new logging LDs need to be created

Checking if preserved data LDs need to be created...
No new preserved data LDs need to be created

Checking if system scheduled tasks need to be created...

Checking if the rights assigned to extended roles need to be updated...
Permission denied

Okay, so you have a cabling issue .. that might cause interesting things ... I would look at getting that fixed.

[Morbo]

Yep, we try to fix it, but more important issue is at bottom

Code: Select all

Checking if the rights assigned to extended roles need to be updated...
Permission denied

i`ve run admithw on second 3par and it went trough without errors and did a lot of work : )
But on this it stops after this permission error.

[MammaGutt]

Yep, we try to fix it, but more important issue is at bottom

Code: Select all

Checking if the rights assigned to extended roles need to be updated...
Permission denied

i`ve run admithw on second 3par and it went trough without errors and did a lot of work : )
But on this it stops after this permission error.

It could be related to cabling... but what user are you using and how are you logging on?

[Morbo]

on both systems i`ve logged on with 3paradm user via direct ssh connection.

[MammaGutt]

on both systems i`ve logged on with 3paradm user via direct ssh connection.

That should be correct. I would try and get the cables sorted out (https://support.hpe.com/hpesc/public/do ... -c03607614 ).

[Morbo]

I`ve manually remove space chunklest from all SSD`s on 3par-1 system to match counts to 3par-2. It freed up space to enable AFC.
admithw still not working properly.
Next thing - we schedule recabling and try again.
Thank you for your support, i`ll back after recabling.