add cmk in KuD

- deploy cmk related pods
- untaint compute nodes if necessary
- run cmk unit tests: allocate CPUs from exclusive and shared pools
- deploy a testing nginx pod along with cmk testing pods
- preset 1/2 CPUs for shared/exlusive pools to fit CI server machines
  users can adjust the parameters to meet their own requirements

Test Results:
  - many rounds of vagrant/5 VMs(controller01/02/03 and compute01/02)
    based test are all OK
  - 14 rounds tests on my local server (S2600WFQ (36C/72T) )and
    PC(HP Z228 (4C/4T)) with all-in-one bare metal deployment are all OK
  - CI(a 4C/4T machine) results of latest patch set also show that the
    test of bare metal deployment is OK
  - NOTE: both my local test and CI use the same testing method of calling
          aio.sh after applying the latest patch set.

Change-Id: I046a4a63b94f92f23347ab76c21a661521e01119
Issue-ID: MULTICLOUD-879
Signed-off-by: Liang Ding <liang.ding@intel.com>

0 files changed, 0 insertions, 0 deletions