A common challenge for OpenStack and K8S deployments is debugging the network when things go awry. The aim of DPHAT is to provide operators of cloud infrastructure with tooling that can analyze the environment and supply the following:
- Feedback that the environment is in a healthy operational state
- Identification of and guidance about where something in the network fabric is broken
- Guidance on remediation steps
- A pluggable interface to enable support for various cloud platforms, their respective networking backends, and any hardware devices (ie switches/routers) present in the deployment
- RESTful API, CLI, and UI
- Gathering information from any relevant SDN controller, representing the network topology for the cloud, and developing an algorithm for analyzing the topology
- Probing of VM's and containers via ARP, ICMP (ping), port scan, ofproto trace, etc. to asses forwarding and security policy instantiation
- Reading pod / compute node state and identifying missing namespaces, tap devices, iptables chains, etc.
- Building a database of remediation actions that can be correlated with issues flagged by DPHAT
If you want to help alleviate the headache of debugging networking issues in the cloud, let's work together!
This project is part of:
Hack Week 18