We often use SnapMirror to move volumes on the same node and also the WFA LUN migrations workflow where we have encountered a problem whcih i belive comes from the way the \etc\Hosts is re-written by the PowerShell cmdlets.
Example using NetApp powershell cmdlets version 3.2:
After adding a route using Add-NaNetRoute, etc\rc & \etc\hosts files are re-written however when the \etc\hosts is re-written, it now associates the node name alias with the cluster interfaces which then causes the SnapMirrors to itself to fail until it is removed.
Please see below sanatised \etc\hosts file from one of our 7-mode systems. Note the Cluster interfaces now have a hostname of "node02" & and alias of "node02-c0a". Node02 is the name of the system. This was not there before running any PowerShell networing cmdlets and is the casue of a few SnapMirror issues for us.