workflow:
- start centos daemon set with net=host, inside pod start python HTTP server
- start centos daemon set with default net and expose it, inside pod start python HTTP server
- with 'kubectl exec' check that python HTTP server is accessible for every pod that we started (for 7 node cluster it starts 14 pods, runs 294 checks and takes 5 minutes if all ports are accessible)
- send telegram msg if needed
- remove daemonsets and services
example usage:
export TELEGRAM_TOKEN=xxx && export TELEGRAM_CHAT=xxx && ./check-network.sh 38743 yes yes