rsevilla87
commented
4 months ago $ ./bin/amd64/k8s-netperf --config examples/netperf-full.yml --debug --all
INFO[2024-04-30 11:35:32] Starting k8s-netperf (141@b939249af9389f95d195eedcb47dce131a209342)
INFO[2024-04-30 11:35:32] 📒 Reading examples/netperf-full.yml file.
INFO[2024-04-30 11:35:32] 📒 Reading examples/netperf-full.yml file - using ConfigV2 Method.
INFO[2024-04-30 11:35:32] Cleaning resources created by k8s-netperf
INFO[2024-04-30 11:35:35] 🔬 prometheus discovered at openshift-monitoring
INFO[2024-04-30 11:35:36] 🔨 Creating namespace: netperf
INFO[2024-04-30 11:35:37] 🔨 Creating service account: netperf
WARN[2024-04-30 11:35:38] ⚠️ Single node per zone and/or no zone labels ]
EBU[2024-04-30 11:35:38] Number of nodes with role worker: 3
DEBU[2024-04-30 11:35:38] Looking for service iperf-service in namespace netperf
INFO[2024-04-30 11:35:38] 🚀 Creating service for iperf-service in namespace netperf
DEBU[2024-04-30 11:35:39] Looking for service uperf-service in namespace netperf
INFO[2024-04-30 11:35:39] 🚀 Creating service for netperf-service in namespace netperf
INFO[2024-04-30 11:35:40] 🚀 Starting Deployment for: client-host in namespace: netperf
INFO[2024-04-30 11:35:40] ⏰ Checking for client-host Pods to become ready...
INFO[2024-04-30 11:35:42] 🚀 Starting Deployment for: client-across in namespace: netperf
INFO[2024-04-30 11:35:42] ⏰ Checking for client-across Pods to become ready...
INFO[2024-04-30 11:35:45] 🚀 Starting Deployment for: server-host in namespace: netperf
INFO[2024-04-30 11:35:46] ⏰ Checking for server-host Pods to become ready...
INFO[2024-04-30 11:35:50] 🚀 Starting Deployment for: server in namespace: netperf
INFO[2024-04-30 11:35:50] ⏰ Checking for server Pods to become ready...
DEBU[2024-04-30 11:35:55] server Running on ip-192-168-218-97.us-west-2.compute.internal with IP 192.168.218.97
DEBU[2024-04-30 11:35:56] client-across Running on ip-192-168-217-224.us-west-2.compute.internal with IP 192.168.217.224 # <----- Host network
DEBU[2024-04-30 11:36:01] Executing workloads. hostNetwork is true, service is false
DEBU[2024-04-30 11:36:01] 🔥 Client (client-host-55b9bd9c4-h9mr5,192.168.218.97) starting netperf against server : 192.168.217.224
INFO[2024-04-30 11:36:01] 🗒️ Running netperf TCP_STREAM (service false) for 5s
DEBU[2024-04-30 11:36:01] [super-netperf 1 42424 -H 192.168.217.224 -l 5 -t TCP_STREAM -- -k
......
DEBU[2024-04-30 11:38:15] Executing workloads. hostNetwork is false, service is false
DEBU[2024-04-30 11:38:15] 🔥 Client (client-across-64f9ddff6b-qmsp4,10.129.1.113) starting netperf against server : 10.130.0.151 # <----- Pod network
INFO[2024-04-30 11:38:15] 🗒️ Running netperf TCP_STREAM (service false) for 5s
DEBU[2024-04-30 11:38:15] [super-netperf 1 42424 -H 10.130.0.151 -l 5 -t TCP_STREAM -- -k rt_latency,p99_latency,throughput,throughput_units,remote_recv_calls,local_send_calls,local_transport_retrans -m 64 -R 1]
DEBU[2024-04-30 11:38:24] MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 42424 AF_INET to 10.130.0.151 () port 42424 AF_INET : demo
RT_LATENCY=-1.000
P99_LATENCY=2
THROUGHPUT=1225.28
LOCAL_TRANSPORT_RETRANS=0
REMOTE_RECV_CALLS=90990
LOCAL_SEND_CALLS=11966562
THROUGHPUT_UNITS=10^6bits/s
Earlier --all flag used to run both podNetwork and hostNetwork tests. However, this was broken recently and this patch addresses this issue.
Type of change
Description
Related Tickets & Documents
Checklist before requesting a review
Testing