Increase connect-timeout in benchmark tests

Currently, while using subctl benchmark tests, the client pod is using a connect-timeout of 10ms which is too low. It was working fine in KIND Clusters as the latency between the clusters is generally less, but in an OnPrem vs Public cluster environments the benchmark tests it was creating issues. This PR fixes the issue by increasing the connect-timeout value to 10 secs. Signed-Off-by: Sridhar Gaddam <[email protected]> (cherry picked from commit 3a6730c)
submariner-io · May 4, 2021 · de2e5b2 · de2e5b2
1 parent d537cfe
commit de2e5b2
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/test/e2e/framework/network_pods.go b/test/e2e/framework/network_pods.go
@@ -345,7 +345,7 @@ func (np *NetworkPod) buildThroughputClientPod() {
 						{Name: "TARGET_PORT", Value: strconv.Itoa(np.Config.Port)},
 						{Name: "CONN_TRIES", Value: strconv.Itoa(int(np.Config.ConnectionAttempts))},
 						{Name: "RETRY_SLEEP", Value: strconv.Itoa(int(np.Config.ConnectionTimeout))},
-						{Name: "CONN_TIMEOUT", Value: strconv.Itoa(int(np.Config.ConnectionTimeout))},
+						{Name: "CONN_TIMEOUT", Value: strconv.Itoa(int(np.Config.ConnectionTimeout*1000))},
 					},
 				},
 			},