These numbers cannot be compared to a hard disk install
and show significant time accessing an rpmdb.
What is different in how you measured?
Note that varying the package used as input is unlikely
to have any significance (there are changes in the file digest
used from MD5 -> SHA256 which is predictably slower).
These numbers cannot be compared to a hard disk install
and show significant time accessing an rpmdb.
What is different in how you measured?
Note that varying the package used as input is unlikely
to have any significance (there are changes in the file digest
used from MD5 -> SHA256 which is predictably slower).