Q: OpenReplicator hot mode hot pull from Clariion

My question is about the requirements for OpenReplicator in hot mode for a reliable full pull from a Clariion.. THe goal is to have enough redundancy to survive cable failures during the hot pull.

For hot mode, it is necessary to have a connection from every Symmetrix FA processor towhih the hot pulled volume is mapped to every SP of the Clariion from which you want to pull. If allcables are working,one connection is sufficient - but  for redundancy, you SHOULD connect both frontend prots of the processor to both SPs.

This works reliably, and iis not afected  if the source LUN on the Clariion is currently trespassed, and it even provides enough redundancy to survive single cable or SFP falures during the hot pull (and this type of redundancy is IMO a prerequisite fpr hot pull, i.e. for taking the host online while the hot pull is still running). However,it results in an awfully large amount of zones in your SAN - e.g. 48 zones for a DMX with 4 FA cards and 4 RDF links (leaving 24 usable frontend ports).for every DMX-to-Clariion relation. My customer has 6 DMXes,  every DMX is connected to 5 Clariions ... yielding about 1500 zones in the SAN only for OpenReplicator, and yielding awfully many frontend logins in the Cariions. and the customer plans to install more DMXes and Clariions (forrtunately, but I may run into limitations of SAN configurations and/or Clariion frontend connectibity).

This configuration was implemented and tested with Clariion flare 19 and 24. The DMX initiator was registered in standard mode, i.e. failobermode 1 etc. Meanwhile, beginning with flare 26, we have ALUA mode in the Clariions. My question is therefore:

- is it possible (and is it a supported configuration) to connect and register  a DMX OpenReplicator   initiator into the Clariion in failover mode 4, i.e.in ALUA mode?

-is my assumption correct I that it is sufficient to connect one FA prot of a processor to SPA, and the other one to SPB - and that that this will give me the same amount of redundancy (i.e.survive a DMX or Clariion frontend cable failure), but need only half the number of zones and frontend logins?

- will Opren Replicator observe the preferred path in ALUA mode, or must I expect that about 50% of all OR pulls will use the "wrong" path - IOW wil ALUA mode induce the risk of adversely affecting the Clariion?'s performance

TIA