Seeing more Latency for Intra-Node-POD compare with Inter-node-POD interface using Uperf Benchmark Operator

[vpuliyal]

I ran uperf using benchmark-operator on my OCP cluster using below configuration and from results I could see that Inter-node(diff worker node) Latency(ms) is better compare with Intra-Node(Same worker node).

############## POD-Interface Results################# TestType Protocol Size Threads inter-node_POD-Interface intra-node_POD-Interface stream tcp 1024 1 3.16 8.71 stream tcp 1024 2 2.04 4.17 stream tcp 1024 4 1.33 1.81 stream tcp 1024 8 1.25 0.91 ############## POD-Interface Results#################

Can you please help on this why we are seeing this issue?

[vpuliyal]

Here I'm attaching my inter and intra log files, xml files. inter-node_POD-Interface_06Mar2023_070700.log intra-node_POD-Interface_06Mar2023_070700.log inter-node_POD-Interface_06Mar2023_070700.yaml.log

[vpuliyal]

#grep Late intra-node_POD-Interface_06Mar2023_070700.log
 - uperf: 95%ile Latency(ms) : 8.713995382006182 -->> nthreads="1"
 - uperf: 95%ile Latency(ms) : 4.165928104607143 -->> nthreads="2"
 - uperf: 95%ile Latency(ms) : 1.810699533429917 -->> nthreads="4"
 - uperf: 95%ile Latency(ms) : 0.905351680341480 -->> nthreads="8"

[vpuliyal]

# grep Lat inter-node_POD-Interface_06Mar2023_070700.log
 - uperf: 95%ile Latency(ms) : 3.1568741385637407  -->> nthreads="1"
 - uperf: 95%ile Latency(ms) : 2.0403174485061992  -->> nthreads="2"
 - uperf: 95%ile Latency(ms) : 1.3250461337101707  -->> nthreads="4"
 - uperf: 95%ile Latency(ms) : 1.2536918885403647  -->> nthreads="8"

[stale[bot]]

Is this still relevant? If so, what is blocking it? Is there anything you can do to help move it forward?

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.