Hello KVM-Autotest users & developers, I want to present a new KVM-Autotest module here: "kvm_iperf". Basically it tests networking functionality, stability and performance of guest OSes. This test is cross-platform -- i.e. it works on both Linux and Windows VMs. I was under development since some time, and now I feel it is mature for a release. The test is dependent on python and KVM-Autotest framework. Basically the module consists of "kvm_iperf.py", and small modifications to "kvm_runtest_2.py" and "kvm_tests.py". You will also need to create a new "misc" subdirectory inside your "kvm_runtest_2/". (# mkdir kvm_runtest_2/misc) And put two files there: 1. "iperf" -- this one must be Linux i586 binary 2. "iperf.exe" -- this one must be Win32 binary optionally third file: 3. "iperf64" -- this one must be Linux x64 binary On Linux platform we could compile on the fly, but I decided not to (for now), because some Linux guests do not have a compiler. Also using i586 binary ensures that it can be used on both i586 and x64 guests. But you can use both, if you modify "kvm_tests.py" accordingly. optional parameters: iperf_duration -- allows to specify long test durations (default = 5 sec) iperf_parallel_threads -- allows to test multiple network threads in parallel (default = 1 thread) Theoretically in future we could support other UNIXes as well (BSD and Solaris). For this you will need only to add respective binaries in "misc/" folder, and modify "kvm_tests.py" for respective OSes. iperf: http://sourceforge.net/projects/iperf - source for latest. iperf for Windows: http://dast.nlanr.net/Projects/Iperf/iperf-1.7.0-win32.exe - stable binary. http://noc.pregi.net/iperf.html -Alexey Eromenko
Attachment:
kvm_iperf.py
Description: Binary data
Attachment:
kvm_runtest_2.py.patch
Description: Binary data
Attachment:
kvm_tests.cfg.sample.2009-04-26
Description: Binary data
Attachment:
kvm_tests.cfg.sample.2009-04-26.iperf.patch
Description: Binary data
Attachment:
kvm_tests.cfg.sample.2009-04-26.iperf.patched
Description: Binary data