Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/remram44/kubernetes-network-checker
Checks network connectivity between pods and services on all hosts
https://github.com/remram44/kubernetes-network-checker
Last synced: 20 days ago
JSON representation
Checks network connectivity between pods and services on all hosts
- Host: GitHub
- URL: https://github.com/remram44/kubernetes-network-checker
- Owner: remram44
- License: bsd-3-clause
- Created: 2022-05-04T15:41:08.000Z (over 2 years ago)
- Default Branch: master
- Last Pushed: 2023-09-27T18:56:32.000Z (about 1 year ago)
- Last Synced: 2024-10-11T07:27:57.369Z (about 1 month ago)
- Language: Python
- Size: 96.7 KB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
Kubernetes Network Checker
==========================
This is a simple tool that will run a pod on each node of your cluster, then try to reach other pods via a service. It then prints a matrix which should immediately show any networking issue.
Example:
```
$ kubernetes-network-checker --once --config ~/.kube/config
2022-05-04 11:32:04,241 INFO kubernetes_network_checker: Discovered 6 nodes
2022-05-04 11:32:04,959 INFO kubernetes_network_checker: Created pods
2022-05-04 11:32:05,503 INFO kubernetes_network_checker: Created services
2022-05-04 11:32:10,743 INFO kubernetes_network_checker: 6 pods started
2022-05-04 11:32:25,067 INFO kubernetes_network_checker: Test complete:
master1 master2 master3 worker4 worker5 worker6
-------- --------- --------- --------- --------- --------- ---------
master1 ok ok ok ok ok
master2 ok ok ok ok ok
master3 ok ok ok ok ok
worker4 ok ok ok ok ok
worker5 ok ok ok ok ok
worker6 ok ok ok ok ok
2022-05-04 11:32:25,297 INFO kubernetes_network_checker: Deleting 6 pods...
2022-05-04 11:32:28,624 INFO kubernetes_network_checker: Deleting 6 services...
```
You can either run the tool on your machine (use `--config`) or run it inside the cluster as a pod (it will use in-cluster configuration, e.g. a service account).
You can either run it once (use `--once`) or let it run continuously, in which case it will run a check every 15 minutes. In this case it will also expose metrics (for Prometheus) which you can use for alerting.