How to get involved
From Linux NFS
The NFSv4 development and testing efforts are organized along the lines of typical Open Source projects. Most discussion occurs on the nfsv4 mailing list, including bug reports, requests for help, test results and so forth.
There is also a bug tracker you can review to see if your bug is already known prior to reporting it on the list. The bug tracker is still fairly new, but hopefully by the time users are incorporating nfsv4 into their environments, many of the most common known bugs will be listed there.
We also have a wiki available for collaborating on information related to testing and troubleshooting of NFSv4.
The community has developed a Test Matrix spreadsheet that itemizes in great detail the testing tasks that should be done, prioritized High/Med/Low. This spreadsheet is being [moved into the http://developer.osdl.org/dev/nfsv4/wiki/index.php/Test_matrix wiki] to make it easier for testers to update.
An easy way to contribute right now, even if you have not started planning your testing efforts, is to write a use case that generally describes your architectural layout and identifies the characteristics of NFS that you think you will be stressing. By sharing your use case with the community, it will help testers better identify areas of NFS to test, which means when you do start doing your NFSv4 testing, you may find it more suitable to your own needs than otherwise.
When you begin doing your NFSv4 testing, we would encourage you to review the test matrix. This may save you some effort if someone has already done the testing, as you may be able to reuse their results or test scripts. It may also help you identify specific things to test for your own needs that no one else has tested. In this case, please list yourself for that item and post links to your results and test scripts, so others can check your findings (especially if you've found problems that the NFSv4 developers should fix).
Another important way to contribute to the testing effort is by providing your testing tools for others to use under an Open Source license. This makes it possible for the NFSv4 developers to try to replicate your issue, and allows NFSv4 testers to use your test to check for regressions in new code. This can save you a lot of headaches if you're particularly dependent on an uncommon feature that doesn't show itself in normal testing activities.
