@Edwints84 What most go with in similar setups is just Active/Standby uplinks (one to each switch) and then the same uplinks being used for vMotion/other storage traffic/replication as Standby/Active so as to avoid contention. The workload of a 3-node cluster (and hence the inter-node traffic) is extremely unlikely to be pushing 10Gbps networking gear to the limits so going with any more complex setup (e.g. link aggregation to try and use multiple uplinks and switches concurrently) shouldn't be necessary nor beneficial.
