root_cmnas01 & root_cmnas02 are the root volumes for the nodes cmnas01 and cmnas02. There is multiple command that can help you validate the upgrade before running it. "cluster image validate -version" and "system node image update.... -estimate-only true" commands.
:*:> cluster image validate -version.........
[-validate-only [true]] - Validate the Package before InstallationUse this parameter to validate the package. Validation consists of verifying whether there is enough space on the system to install the package, verifying the checksum for each component within the package and so on. Validation usually takes from 30 to 60 minutes. If you specify this parameter, the package will be validated only, not installed.
http://docs.netapp.com/ontap-9/topic/com.netapp.doc.dot-cm-cmpr-920/cluster__image__validate.html
::*> system node image update....
[-estimate-only [true]] - Estimate OnlyCreates a report of the steps that occur during the update without actually doing them.https://docs.netapp.com/ontap-9/topic/com.netapp.doc.dot-cm-cmpr-920/cluster__image__update.html
That been said, you can always go to systemshell and clear /etc/crash directory. Old core files and packet traces can be taking space from the root volumes.
Regarding the cm_svm_root volume. this is the root volume for the vserver. There is no need to access that volume,