As for "standard", I'm not sure there is one. What I think you might want to do is create a set of tests that are standard for you and apply that to your variations.
I still continue to use the old^H^H^Hclassic standbys for disk performance testing: Iozone and iometer. Iozone is fairly easy to get going, while Iometer can be configured to more closely match your real-life load characteristics (assuming you know what they are, of course). I believe both are available on multiple platforms, including Linux and Windows.
There are many tools out there that perform stress testing.
For Linux a easy to use one is stress: http://weather.ou.edu/~apw/projects/stress/
for the disk: bonnie++. not your most trendy benchmarking tool but one that gives useful results [ random and sequential reads/writes, i/o per sec ].
One test I do is compile the kernel. Maybe prime. Also I'm refering to this list quite a lot lately as I am testing a lot of VM configurations.
http://www.overclockers.com/forums/showthread.php?t=486495
As for "standard", I'm not sure there is one. What I think you might want to do is create a set of tests that are standard for you and apply that to your variations.
I still continue to use the old^H^H^Hclassic standbys for disk performance testing: Iozone and iometer. Iozone is fairly easy to get going, while Iometer can be configured to more closely match your real-life load characteristics (assuming you know what they are, of course). I believe both are available on multiple platforms, including Linux and Windows.