mrahbar/k8s-testsuite

Test suite for Kubernetes

This test suite consists of two Helm charts for network bandwith testing and load testing a Kuberntes cluster. The structure of the included charts looks a follows:

k8s-testsuite/
- load-test/
|____Chart.yaml
|____values.yaml
|____.helmignore
|____templates/
- network-test/
|____Chart.yaml
|____values.yaml
|____.helmignore
|____templates/

You can install a test by running:

> git clone https://github.com/mrahbar/k8s-testsuite.git
> cd k8s-testsuite
> helm install --namespace load-test ./load-test
> helm install --namespace network-test ./network-test

Load test

This Helm chart deployes a full load test suite in Kubernetes. It consists of the 3 microservices:

1. A webserver based on simple-webserver
2. A loadbot client which is based on this and that
3. An aggregator which orchestrates the test run

Aggregator

Since the webservers and the loadbots work autonomic the task of the aggregator ist to orchestrate the test run. It does this by useing the Kubernetes api via client-go library to talk set up the desires replicas of each unit. The test run consists of the following tests scenarios:

The maximum count of replicas (default 100) can be set with --set aggregator.maxReplicas=....

Loadbots

The loadbots have the task to run a predefined level of queries per second. Vegeta publishes detailed statistics which will be fetched and evaluated by the aggregator. This metrics are:

• Queries-Per-Second (QPS)
• Success-Rate
• Mean latency
• 99th percentile latency

Test results

When all tests are finishes the aggregator will print the following summary to its logs:

GENERATING SUMMARY OUTPUT Summary of load scenarios: 0. Idle : QPS: 10037 Success: 100.00 % Latency: 949.82µs (mean) 3.004154ms (99th)

1. Under load: QPS: 10014 Success: 100.00 % Latency: 965.549µs (mean) 1.985838ms (99th)
2. Equal load: QPS: 50078 Success: 100.00 % Latency: 982.519µs (mean) 7.213018ms (99th)
3. Over load : QPS: 501302 Success: 100.00 % Latency: 198.21451ms (mean) 859.504601ms (99th)
4. High load : QPS: 502471 Success: 100.00 % Latency: 239.26364ms (mean) 1.018523444s (99th) END SUMMARY DATA

Configuration

The following table lists the configurable parameters of the chart and their default values.

Network test

This Helm chart deployes a network test suite in Kubernetes. It consists of the 2 microservices:

1. An orchestrator
2. A worker launched three times

Both services are run from the same image either as --mode=orchestrator or --mode=worker. The services are bases on k8s-nptest and use iperf3 and netperf-2.7.0 internally.

Orchestrator

The orchestrator pod coordinates the worker pods to run tests in serial order for the 4 scenarios described below. Using pod affinity rules Worker Pods 1 and 2 are placed on the same Kubernetes node, and Worker Pod 3 is placed on a different node. The nodes all communicate with the orchestrator pod service using simple golang rpcs and request work items. A minimum of two Kubernetes worker nodes are necessary for this test.

Test scenario

Five major network traffic paths are combination of Pod IP vs Virtual IP and whether the pods are co-located on the same node versus a remotely located pod.

1. Same VM using Pod IP: Same VM Pod to Pod traffic tests from Worker 1 to Worker 2 using its Pod IP.

2. Same VM using Cluster/Virtual IP: Same VM Pod to Pod traffic tests from Worker 1 to Worker 2 using its Service IP (also known as its Cluster IP or Virtual IP).

3. Remote VM using Pod IP: Worker 3 to Worker 2 traffic tests using Worker 2 Pod IP.

4. Remote VM using Cluster/Virtual IP: Worker 3 to Worker 2 traffic tests using Worker 2 Cluster/Virtual IP.

5. Same VM Pod Hairpin: Worker 2 to itself using Cluster IP

For each test the MTU (MSS tuning for TCP and direct packet size tuning for UDP) will be linearly increased from 96 to 1460 in steps of 64.

Output Raw CSV data

The orchestrator and worker pods run independently of the initiator script, with the orchestrator pod sending work items to workers till the testcase schedule is complete. The iperf output (both TPC and UDP modes) and the netperf TCP output from all worker nodes is uploaded to the orchestrator pod where it is filtered and the results are written to the output file as well as to stdout log. Default file locations are /tmp/result.csv and /tmp/output.txt for the raw results.

All units in the csv file are in Gbits/second

ALL TESTCASES AND MSS RANGES COMPLETE - GENERATING CSV OUTPUT
the output for each MSS testpoint is a single value in Gbits/sec 
MSS , Maximum, 96, 160, 224, 288, 352, 416, 480, 544, 608, 672, 736, 800, 864, 928, 992, 1056, 1120, 1184, 1248, 1312, 1376, 1460
1 iperf TCP. Same VM using Pod IP ,24252.000000,22650,23224,24101,23724,23532,23092,23431,24102,24072,23431,23871,23897,23275,23146,23535,24252,23662,22133,,23514,23796,24008,
2 iperf TCP. Same VM using Virtual IP ,26052.000000,26052,0,25382,23702,0,22703,22549,0,23085,22074,0,22366,23516,0,23059,22991,0,23231,22603,0,23255,23605,
3 iperf TCP. Remote VM using Pod IP ,910.000000,239,426,550,663,708,742,769,792,811,825,838,849,859,866,874,883,888,894,898,903,907,910,
4 iperf TCP. Remote VM using Virtual IP ,906.000000,0,434,546,0,708,744,0,791,811,0,837,849,0,868,875,0,888,892,0,903,906,0,
5 iperf TCP. Hairpin Pod to own Virtual IP ,23493.000000,22798,21629,0,22159,21132,0,22900,21816,0,21775,21425,0,22172,21611,21869,22865,22003,22562,23493,22684,217872,
6 iperf UDP. Same VM using Pod IP ,6647.000000,6647,
7 iperf UDP. Same VM using Virtual IP ,6554.000000,6554,
8 iperf UDP. Remote VM using Pod IP ,1877.000000,1877,
9 iperf UDP. Remote VM using Virtual IP ,1695.000000,1695,
10 netperf. Same VM using Pod IP ,7003.430000,7003.43,
11 netperf. Same VM using Virtual IP ,0.000000,0.00,
12 netperf. Remote VM using Pod IP ,908.460000,908.46,
13 netperf. Remote VM using Virtual IP ,0.000000,0.00,
END CSV DATA

Configuration

The following table lists the configurable parameters of the chart and their default values.