In one of the previous blog post I showed the usage of rook operator to install ceph on top of kubernetes. I have followed the same link to create ceph cluster on kubernetes and verified that PVC is getting created successfully.

In one of the previous article I talked about the Kubernetes monitoring. In this article I am covering the deployment of prometheus operator on kubernetes and steps to start monitoring your application.

In case of kubernetes we can scale the application based on CPU and memory utilization but sometimes this is not sufficient. We need other parameters specifically related to application for scaling the number of PODs. Fortunately, kubernetes prometheus has another project adapter which can be used to expose the metrics from prometheus which then can be use for scaling the app.

In this article, I am providing the steps to add new alerting rules for an app in prometheus. Once you have deployed the prometheus using prometheus operator then you can login into the prometheus container to understand the configuration. It’s showing us that file /etc/prometheus/config_out/prometheus.env.yaml is used as prometheus configuration file.

configmap is very important topic in kubernetes, this is a recommended way to pass the configuration to your PODs instead of hard-coding the config files into your docker images. In this article, I provide the information about various ways of providing configmap to PODs.

For a long time I was looking for a way to run the multiple minikube VMs on a single machine for testing out federation working. I came to know about the profile feature of minikube and eagerly tried but it was failing for me. I filed a bug for this issue.

It’s very common to have the requirement to change the parameters for a kubernetes service. I was in need of testing the PodSecurityPloicy but I found that it’s not present by-default in the --admission-control parameter list of kube-api service.

Recently I deployed multinode HA setup using kubespray, after that I was able to issue the commands from the master node but was not able to issue the commands from my laptop. I am using my same laptop for creating the minikube setup also and I knew that kubernetes has concept of context which is used to access the K8 setups.

Recently I was working on simple app which I wanted to deploy in K8, it was cumbersome task to write some code and then doing everything manually to ensure that it’s working fine in K8 setup. I came across a wonderful tool ksync which helps to sync the changes which are doing locally directly in the K8 POD. This works in client/server architecture similar to helm. Once you dowmnload the kysnc initialize the ksync again similar to helm, it will start daemonset in the K8.

Recently I deployed the multinode K8 setup using kops on my Mac with vagran and virtual box as a driver. It was very easy to complete this setup I only issued one command.