After a panic last February all disk ownerships are reassign so they no longer depending on the disks stack membership. Originally, all disk of one stack were belonging to the same controller. Is their any reason to be concerned about this new "chaos"?
A short description of the environment: HA pair fas3270, 2 stack with 2 ds4243 each, multipath sas cabling and acp cabling, originally all disks of stack A were belonging to controller A and these of stack B to the partner controller B. Now all disk ownerships are distributed randomly over these 2 stacks and contollers (SW based disk ownership)
Ignoring issues relating to ease of maintenance i think: Technically there is no important difference either disk ownership is "stack-straight" or not. But i did not find any clear statement in netapp documentation confirming this statement. Any hints?
When you assign ownership for disks, follow these guidelines to keep autoassignment working and to maximize fault isolation.
Always assign all disks on the same loop or stack to the same system and pool.
Always assign all loops or stacks connected to the same adapter to the same pool.
Note: You can configure your system to have both pools on a single loop or stack. On storage system models that only support one loop or stack, this configuration cannot be avoided. However, in this configuration, a shelf failure would cause a data service outage.