MSA 2050 - Pool A & B

[Alessandro_Del]

Dear all, I need your support to understand some things on my new SAN HPE 2050.

given: SAN ISCSI 10GB - advanced license - 3 node Hyper-V cluster

By configuring disk groups I have the possibility to assign arrays to pool A or pool B
I bought 12 disks (6 sas 15K + 6 ssd)
My project was:
group disk 1 - raid 5 - 5 ssd
group disk 2 - global spare - 1 ssd
group disk 3 - raid 5 - 5 sas 15k
group disk 4 - global spare - 1 sas 15k
1 or 2 logical volumes on disk group (A or B or bothA + B? this is now my question).

do you have to do something for failover or are both pool A and pool B already clustered on storage processors??
So can I create everything on A or do I have to split between pool A and pool B?

For the purposes of the mapping I see that only 4 ports are seen (all with the check enabled) I assume that they are clustered on both controllers (I have 4 ports for each controller).
Don't I have to change anything?
It's correct?

I appreciate suggestions, thanks.

[SUBHAJIT KHANBARMAN_1]

@Alessandro_Del 

You need to purchase HPE MSA Advanced Data Services Suite License.

You can create SAS Virtual Disk Group(VDG) and SSD Virtual Disk Group(VDG) as part of same Pool so that you can take advantage of Tiering concept.

Now you can create Voluemes out of that Pool and present the volume to host with all host ports both through Controller A and Controller B. This will help to provide redundancy as well as number of paths will increase.

In case Volume owning controller failed or down then other Controller will take care automatically.

Hope this helps!
Regards
Subhajit

I am an HPE employee

If you feel this was helpful please click the KUDOS! thumb below!

***********************************************************************

[Alessandro_Del]

Hello Subhajit, 

thanks a lot for your reply.

I have already activated Advanced Data Services Suite License correctly.

My doubt was precisely about assigning the disk group to pool A or B.
If you can clarify the concept of pool A and B I would be grateful.
Apart from that, I can proceed indifferently on pool A or B and then create the volume or volumes as needed. (about this point, is it recommended to create a single volume or 2 volume? )

About the controller ports, is it correct that I only see 4 ports? are they already aggregated?
That's where I got confused thinking that 4 ports are for pool A and then controller A and 4 ports for pool B and then controller B.

Many thanks for your support, 

BR

Alessandro

[SUBHAJIT KHANBARMAN_1]

@Alessandro_Del 

Pool A is nothing but Virtual Pool A which is owning by Controller A

Pool B is nothing but Virtual Pool B which is owning by Controller B

Regarding Volume creation, it depends upon your business requirement. For example, if you have two departments where you need to give data access separately then better to create two volumes.

Now volume should be part of same Pool or not, regarding this I would say better to be part of same pool so that you can take advantage of tiering.

There is no port aggregation concept here. Each host port works individually. So per controller if you see 4 ports then it means 4 separate paths. As explained last time always better to present volumes with both controller host ports in order to achieve redundancy.

Hope this helps!
Regards
Subhajit

I am an HPE employee

If you feel this was helpful please click the KUDOS! thumb below!

***********************************************************************

[Alessandro_Del]

Hello

Thanks again,

I can conclude by saying that if virtual pool A has all the disks in charge and on pool B there is nothing, in case of controller A fail, the virtual pool A with the disks and volumes, is managed automatically by controller B .

Just make sure that all network ports are mapped to the volume and to the host.

Best regards,

Alessandro.

[SUBHAJIT KHANBARMAN_1 · Accepted solution]

@Alessandro_Del 

Yes your understanding is correct

Do you have any outstanding query on this topic?

If you have no further query and you are satisfied with the answer then kindly mark the forum as resolved so that everyone who are following this forum will get the update.

Hope this helps!
Regards
Subhajit

I am an HPE employee

If you feel this was helpful please click the KUDOS! thumb below!

*********************************************************************

[Alessandro_Del]

Last clarification on the mapping and I can close the topic with great pleasure and satisfaction for your support.

When I make volume association to the host, I only read 4 ports which are automatically selected.
I have connected port A1 10GB and port B1 10GB to my switch.
Can I consider that the 4 selected ports are A1 + B1 - A2 + B2 etc ... or they are only A1 - A2 - A3 - A4 ??
In the first case, it is OK I have nothing to do otherwise, if the ports are only fo controller A, what should I do to present the ports of also controller B?

Can I be sure that if A1 cable is unplugged all traffic goes on B1 automatically?

Thanks.

Alessandro

[Alessandro_Del]

I found answer to my question on best practices.

" Mapping a port makes a mapping to each controller; thus, mapping Port 1 will map host Ports A1 and B1. Mapping to Port 2 will map host Ports A2 and B2."

it also suggested to have both controllers A and B working for a balance workload.
Thanks for support.

Alessandro