This does not directly answer your question but solves the issue of freeing up resources without risk of damaging what is currently running live.
Instead of upgrading in place, if possible lay down a new UNIX server with a scratch install to the highest OS version that your new cloverleaf will run on.
Next install the more recent version of cloverleaf on the new UNIX server.
When ready to run hcirootcopy to upgrade your old sites, do a NFS
Then create a symbolic link to it so hcirootcopy thinks it co-exists on the new UNIX server.
For example we upgraded from Cloverleaf 5.6 to 6.0 so my symbolic link to fool hcirootcopy looks like this:
> pwd
/cloverleaf
> ls -l
total 0
drwxrwxr-x 3 hci staff 256 Feb 19 2013 cis6.0
lrwxrwxrwx 1 hci staff 41 Mar 1 2013 qdx5.6 -> /nfs/mdahub8/cloverleaf/qdx5.6
If successful, at this point while on the new UNIX server you do
cd /cloverleaf/5.6; ls -l
then you will see all the cloverleaf 5.6 sites form the old server but it appears they exist on the new server, which is good enough for hcirootcopy.
Now I can do a hcirootcopy of each old cloverleaf site without having any risk of impact to them since they aren’t even running on my new cloverleaf server.
This approach also allows you to upgrade all aspects of the UNIX server to the highest level while leaving the older server intact untouched.
Since you are doing a clean scratch install the issue of unnecessary wasted space is no longer an item of concern because that goes away if and when the old server is reclaimed.
Russ Ross
RussRoss318@gmail.com
.
