Unsolved

2 Posts

895

November 2nd, 2020 09:00

DELL ME4084 - Using/deleting existing volumes for another initiator

Hi guys.

We currently have a ME4084, 70 disks (SAS),  2 disk groups (35 disk each), 2 controllers, 2 pools (maximum), with 10 volumes total (5 mapped to pool A, and 5 to pool B). Each volume is 61.9TB in size.

As far as the PV admin guide says:

"You can remove volumes from a volume group. You cannot remove all volumes from a group. At least one volume must remain. Removing a volume from a volume group will ungroup the volumes but will not delete them"

Understood, I wanted to see if I could reduce the volume size but:

  "you can expand the size of the volume but not make it smaller"

Our situation:

  We have ~600TB of total capacity. Those are mapped already (and being used) to a Windows server initiator (iSCSI). The ME4084 documents are not really clear and my explanation and question is:

If I need to take ~100TB (let's say I want to use 2 volumes of the current 10, leaving 8 volumes for the windows server host) and then I want to use those two volumes to create an initiator and map the volumes to another host (Linux's Open-iSCSI initiator). I know how to do it, but I don't know how the volume creation or volume deletion will affect the current data (my main concern)

I've already tried to create volumes, but the GUI in the web interface show those "outside the pools" (to the right, like this 50TB (for each pool) are overcommitted) and not inside the pools (I didn't pressed OK as I was just observing the distribution of those in the pool, but they were outside),  I was trying to allocate 50TB on one pool and 50TB on another (2 volumes).

I believe I might have to remove 2 volumes and this is my issue with the documentation: It doesn't say if the current data (if any) on those volumes will be moved over the rest of the disks. I assume that I/O will need to be stopped but still not sure if by removing a volume the data is migrated.

We currently have two 310TB pools and for each ~54TB is being actually allocated for a total of ~100TB

Please advice on how the appliance would behave in terms of the data in the volumes.

 

 

November 2nd, 2020 14:00

Hello,

 

When you create a disk group it is assigned to a pool and assigned to a controller. The volumes (virtual disks) and the data will reside within that disk group. Once they are created they cannot be moved from one pool to the or diskgroup to another. Your data will need to be copied or moved (thru the host OS) from the volume being removed to an existing volume.

 

Please let me know if you have additional questions.

 

1 Message

November 2nd, 2020 20:00

I have exactly the same problem. Anyone got a suggestion?

November 3rd, 2020 06:00

Good morning,

 

I would suggest migrating your data using the host OS from it's current location to an existing location. When you create a disk group it is assigned to a pool and assigned to a controller. The volumes (virtual disks) and the data will reside within that disk group. Once they are created they cannot be moved from one pool to the or diskgroup to another.  

2 Posts

November 4th, 2020 04:00

How many disk failures ADAPT can survive? In the manual in one table it mentions two disks.

November 4th, 2020 07:00

Hello,

 

Like you saw in the manual, 2.

Top