- a test infrastructure.
- designed to allow an easy start.
- designed to extend to complex test labs
- able to unify lots of different use-cases under a common umbrella
- generic enough to compete with the future
at AMD's Operating System Research Center where it served as umbrella for all QA. It integrates 10+ years of QA know-how and Linux testing expertise since 2008. As of 2014 it is in active use and development at AWS. It also provides the technology behind the Perl benchmarking infrastructure Perl::Formance.
- extreme continuous integration through setting up complete machines from scratch, either image-based or via kickstart/autoyast/preseed.
- alternative easy local or ssh-based test invocation
- test facilities for the Linux kernel, virtualization based on Xen and KVM, and running in simulators (like simnow)
- support for lots of other scenarios, like suspend/resume or distributed network tests
- benchmark tracking facilities
- powerful, advanced query language for test result evaluation and forensics
- a powerful scheduler to maximize utilization of a machine pool with different use-cases organized in bandwidth-driven queues
- are easy, without requiring dependencies to talk to them
- allow integration of your existing infrastructure with nothing more like shell scripts using "echo" and "netcat"
the use of language diversity for writing tests by building upon TAP, the mature "Test Anything Protocol", supported in about 20 programming languages.
- any TAP emitters