I work mainly in server/storage, but 40 Gbps is picking up for NFS access to storage systems. A colleague has tried to get to the bottom of this, since from my understanding 40 Gbps is actually 4x10 Gbps.
Is this "teaming" or "bundling" completely transparent to the TCP/IP layer, or do we actually get 4 sessions of 10 Gbps? A single hypervisor may connect to a storage array, and when using LACP we had to create multiple logical interfaces to get the full advantage of using all links. Perhaps a noob question, but I've not been able to get a solid answer.
No comments:
Post a Comment