Jose Barreto wrote a blog about my favorite changes in Windows Server 2012 R2.
Automatic SMB Scale-Out Rebalancing. Why is it cool? Easy to answer, when you use a SoFS with Windows Server 2012 and one nodes fail the CSV switches over to the other Clusternode. That is ok but becomes worst when the other Clusternode comes back, because with Windows Server 2012 the is no automated rebalacing or fallback and the Clusternode redirects all traffic throw the CSV owner. This has an enormously performance decrease.
With 2012 R2, this is history and how it works you can read on Jose’s blog. Click here.