not hard just need to schedule downtime. You will need to remove interface from your CIFS server, delete interface, create new device, re-assign interface to the trunk device, re-assign interface back to the CIFS server.

Thanks dynamox,

what about current CIFS shares, do I need to delete them before remove the interface or no need to do anything and the shares will remain after re-assign new interface back to the CIFS server?

ok, great!

No, we are not using Replicator.

Do you recommend link aggregation with more than these 2 devices?

We are planning to implement RecoverPoint in a few moths, any advice?

Thanks.

no need to mess with shares, permissions ..etc, everything will remain intact. Are you using Replicator ?

RecoverPoint is for block replication, it does not use network interfaces on your datamovers.

Link aggregation - what kind of switching environment do you have ? For example on Cisco Nexus switches that use vPC you can configure a trunk interface on VNX where members are connected to two separate Nexus switches. This gives you more throughput and redundancy. Are you using 10g NICs ?

Nexus ? If not, is the same network/VLAN available on both switches ?

We have 2 Cisco Switches 1Gb ports.

Not Nexus and yes, the same network/VLAN is available on both switches.

it's your call but i would do this :

• if you only have two interfaces to work with, do not trunk. Instead connect one interface to switch 1 and connect second interface to switch 2. Then you will create an FSN (fail safe network) device, and after that you would assign interfaces to it.  Yes, this option does not give you 2 x 1g lane "highway" but you can redundancy (one interface is active and one is stadby). If one of those network switches locks up, if one of your network cables get cut, if there is a link down event ..the stanby interface connected to the other switch takes over. Super fast failover, within 2 seconds.

Actually, I have 4 interfaces on each DM and wanted to take advantage from it, not forgetting future needs that may arise as replication. Does this make sense to you or I should use all 4 DM ports for shares ?