Need zoning advise (2 Brocade,Blade,MSA)

[welldone777]

We very need your advise.
See attachment please.
Desired reliability:
1. One Brocade Failed
2. One Eva Controller Failed
3. One optical patch-cord failed
4. One SFP failed

Blade 1 must see EVA and MSA1000 and has:
- 2 VC-FC
- BL460 servers, Qlogic mezanine (will be used VmWare ESX and Win2K3) and Emulex (Win2K3)

Blade 2 has VC-FC (first ports of servers HBA on which). Servers, mezanines and OS - the same.
No availability because no VC-FC.

Can you advice optimal zoning?

Thank you.

[Uwe Zessin]

What I see as 'optimal' zoning:

WWN-based zoning: Put the port WWNs of an initiator port (server's WWN) and the target port(s) of a storage array in each zone. Create separate zones for each server/ storage array.

This (should be) completely independend on which component failed. With zoning you're not really removing paths [except in some special situations], but isolate active components from each other.

[Sivakumar MJ._1]

Check the doc..

[IBaltay]

Hi,

vc-fc configuration

the good idea of your conceipt is to have 2 vc-fc modules (redundancy). The second good idea is to switch on the vc-fc to assign the virtual PWWNs to have no problem during the failure/replacement of the whole host neither in zoning, nor on the MSA host group HBA port assignment. The 3rd good idea is to switch on the dynamic loadbalancing on the vc-fc level.

zoning
needs to be PWWN zonning (soft zoning because of the NPIV (vc-fc) to F-port (Brocades) maping to be able to zone each host separately

resume
Thus you can survive the
1. One Brocade Failure
2. One Eva Controller Failed
3. One optical patch-cord failed
4. One SFP failed