Table of Contents


This is a cross post from The Cloud Report. Be sure to checkout the The Cloud Report website.


What is Ancientt

Ancientt is a tool to automate network testing tools, like iperf3, in dynamic environments such as Kubernetes, Ansible and more to come dynamic environments.

Meaning that, if you want to test the network throughput with iperf3 of your Kubernetes cluster, you can easily do that with Ancientt. A simple “test definitions” file is used to tell Ancientt what exactly to do:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
version: '0'
runner:
  name: kubernetes
  kubernetes: {}
tests:
- name: iperf3-one-random-to-all-nodes
  type: iperf3
  outputs:
  - name: csv
    csv:
      filePath: .
      namePattern: 'ancientt-results-{{ .Data.Tester }}.csv'
  hosts:
    servers:
    - name: one-random
      random: true
      count: 1
    clients:
    - name: all-hosts
      all: true
  iperf3:
    duration: 30

The above “test definitions” file tells Ancinett, to use the Kubernetes “runner”, to run a test from one randomly chosen Node to all the other Nodes (+ itself) in the Kubernetes cluster. Test results are stored as CSV in files named after the given namePattern to make it easy to split results by “server”, “client” and other available variables.

NOTE It doesn’t have to be a Kubernetes cluster, an OpenShift cluster works as well.

Besides being able to run on Kubernetes and OpenShift, you can also provide an Ansible inventory file which also can be used to run tests on it. The targeted hosts need to have the network testing tools installed as of right now. An example Ansible playbook can be found here: GitHub cloudical-io/ancientt - examples/runners/ansible directory.

Why did we create Ancientt

More and more people are moving to platforms like Kubernetes, OpenStack, OpenShift and others. The biggest problem we see in such installments is most of the time the network. Either because of performance, latency or general connectivity issues.

Ancientt right now allows us to easily run iperf3 tests across whole Kubernetes and OpenShift clusters, and normal machines (through Ansible). Doing such tests manually takes a lot of time away from other potentially critical points that should be investigated. With Ancientt you can easily run such tests faster. This should save you time and the results hopefully get you faster to a resolution of the problem and / or improving your network performance.

We’ll hopefully soon be able to add more network testing tools to the list, to make it even easier to run tests. E.g., smokeping, siege and others, should be able to help with latency and performance tests.

Future of Ancientt

For the near future, we’ll focus on making it possible to get the results into Prometheus. This would allow for constant network testing, monitoring and alerting on issues. If you have a good idea on how this can be achieved, feel free to comment on GitHub cloudical-io/ancientt - Issue ‘outputs: Prometheus’ #38.

As written in the Why did we create Ancientt section, we’ll look into adding more network testing tools. For a list of potential new network testing tools to be worked on for Ancinett, checkout GitHub cloudical-io/ancientt Issues with Label `testers.

Wanna try out Ancientt?

Checkout the Ancientt project on GitHub: GitHub cloudical-io/ancientt.

To get you started on the test definition options available, there is a config examples doc page. If you want to dive into all available test defintion options, see onfig structure doc page.

All available documentation can be found in the GitHub cloudical-io/ancientt - /docs directory.

Summary

Ancientt can help you run iperf3 tests, soon other network test tools as well, in a Kubernetes and OpenShift clusters, and using Ansible in an organized and quick way. Most importantly also collect the test results in an orderly fashion in output formats, like CSV, Excel, MySQL and SQLite.

Be sure to check out the project on GitHub: GitHub cloudical-io/ancientt.

We are happy about any contributions, be it documentation or code changes.

If you like the project, be sure to leave a Star to help the project reach more people.

Have Fun!