Skip to content →

Month: April 2017

Let’s run Homer on Kubernetes!

I have to say that Homer is a favorite of mine. Homer is VoIP analysis & monitoring – on steroids. Not only has it saved my keister a number of times when troubleshooting VoIP platforms, but it has an awesome (and helpful) open source community. In my opinion – it should be an integral part of your devops plan if you’re deploying VoIP apps (really important to have visibility of your… Ops!). Leif and I are using Homer as part of our (still WIP) vnf-asterisk demo VNF (virtualized network function). We want to get it all running in OpenShift & Kubernetes. Our goal for this walk-through is to get Homer up and running on Kubernetes, and generate some traffic using HEPgen.js, and then view it on the Homer Web UI. So – why postpone joy? Let’s use homer-docker to go ahead and get Homer up and running on Kubernetes.

Continue reading Let’s run Homer on Kubernetes!

Comments closed

How-to use GlusterFS to back persistent volumes in Kubernetes

A mountain I keep walking around instead of climbing in my Kubernetes lab is storing persistent data, I kept avoiding it. Sure – in a lab, I can just throw it all out most of the time. But, what about when we really need it? I decided I would use GlusterFS to back my persistent volumes and I’ve got to say… My experience with GlusterFS was great, I really enjoyed using it, and it seems rather resilient – and best of all? It was pretty easy to get going and to operate. Today we’ll spin up a Kubernetes cluster using my kube-centos-ansible playbooks, and use some newly included plays that also setup a GlusterFS cluster. With that in hand, our goal will be to setup the persistent volumes and claims to those volumes, and we’ll spin up a MariaDB pod that stores data in a persistent volume, important data that we want to keep – so we’ll make some data about Vermont beer as it’s very very important.

Continue reading How-to use GlusterFS to back persistent volumes in Kubernetes

Comments closed

koko – Connect Containers together with virtual ethernet connections

Let’s dig into koko created by Tomofumi Hayashi. koko (the project’s namesake comes from “COntainer COnnector”) is a utility written in Go that gives us a way to connect containers together with “veth” (virtual ethernet) devices – a feature available in the Linux kernel. This allows us to specify interfaces that the containers use and link them together – all without using Linux bridges. koko has become a cornerstone of the zebra-pen project, an effort I’m involved in to analyze gaps in containerized NFV workloads, specifically it routes traffic using Quagga, and we setup all the interfaces using koko. The project really took a turn for the better when Tomo came up with koko and we implemented it in zebra-pen. Ready to see koko in action? Let’s jump in the pool!

Continue reading koko – Connect Containers together with virtual ethernet connections

Comments closed