NSB NFVi PROX VPE losing many packets 85/58685/8
authorDanielMartinBuckley <daniel.m.buckley@intel.com>
Mon, 18 Jun 2018 18:27:41 +0000 (19:27 +0100)
committerEmma Foley <emma.l.foley@intel.com>
Wed, 27 Jun 2018 11:26:14 +0000 (12:26 +0100)
commit39b3d9ae8ff19a0d64bff0e874f05b6dbcc30b01
tree8235f17b241aefefbd05a1a74aa9a0f93c0b69a9
parent8d3eedf007d338a032d106842cb66baa3a8eb8b7
NSB NFVi PROX VPE losing many packets

JIRA: YARDSTICK-1105

Running a vPE test ends up with up to 25% drooped packets.
This is due to tolerated loss being incorrectly configured,
test duration was too short and test interval was too
short.

Note This is due to the fact that vPE use case have by
default a tolerated loss of 100%, Now changed to 0.001,
same as L2FWD

Also: Grafan dashboard has been updated to show all SUT
CPUS Utilization in 1 panel. And LINE rate MIN, MAX & TEST
Rate have been added

Change-Id: I7adae2199b3f656fe460705b6aeb3aa69c767d50
Signed-off-by: Daniel Martin Buckley <daniel.m.buckley@intel.com>
dashboard/Prox_BM_vPE-4Port-1507804058631.json [deleted file]
dashboard/Prox_BM_vpe-4Port-1529345833067.json [new file with mode: 0644]
samples/vnf_samples/nsut/prox/tc_prox_baremetal_vpe-4.yaml
samples/vnf_samples/nsut/prox/tc_prox_heat_context_vpe-4.yaml
samples/vnf_samples/traffic_profiles/prox_vpe.yaml
yardstick/network_services/traffic_profile/prox_binsearch.py