Fix PROX throughput result calculation 59/57359/8
authorAbhijit Sinha <abhijit.sinha@intel.com>
Mon, 14 May 2018 13:31:52 +0000 (14:31 +0100)
committerAbhijit Sinha <abhijit.sinha@intel.com>
Tue, 15 May 2018 13:33:51 +0000 (14:33 +0100)
The success criteria needed a fix for throughput
calculation. The results on success criteria were
muliplied by 1000*1000 which is not correct and
not needed as this was leading to false results
in grafana.

JIRA: YARDSTICK-1172

Change-Id: I56b24a700f8a565db80897aeab796e7039cae5c9
Signed-off-by: Abhijit Sinha <abhijit.sinha@intel.com>
tests/unit/network_services/traffic_profile/test_prox_binsearch.py
yardstick/network_services/traffic_profile/prox_binsearch.py

index 1b4189b..28840ef 100644 (file)
@@ -68,7 +68,7 @@ class TestProxBinSearchProfile(unittest.TestCase):
 
         # Result Samples inc theor_max
         result_tuple = {"Result_Actual_throughput": 7.5e-07,
-                        "Result_theor_max_throughput": 0.00012340000000000002,
+                        "Result_theor_max_throughput": 1.234e-10,
                         "Result_pktSize": 200}
         profile.queue.put.assert_called_with(result_tuple)
 
index c3277fb..7a84f13 100644 (file)
@@ -116,7 +116,6 @@ class ProxBinSearchProfile(ProxProfile):
                 self.current_lower = test_value
                 successful_pkt_loss = result.pkt_loss
                 samples = result.get_samples(pkt_size, successful_pkt_loss, port_samples)
-                samples["TxThroughput"] = samples["TxThroughput"] * 1000 * 1000
 
                 # store results with success tag in influxdb
                 success_samples = {'Success_' + key: value for key, value in samples.items()}