Taints and Tolerations Example in Kubernetes

-

Taints and Tolerations are advance scheduling mechanism designed to repel pods if they do not have matching toleration. 

Pods are only accepted by Nodes if they match desired toleration otherwise Pods remains in Pending state or look for another nodes.

In today's example , we will configure nodes and pods with taints and toleration respectively. We will make use of Minikube cluster for managing pods.

Pre-requisite : Minikube Cluster 


As you can see in above picture , we have two pod one with toleration defined and another without toleration. Since minikube is just one node cluster , we will use master node and apply taints on it.

We will see pod scheduling behaviour when toleration applied versus toleration not applied. Let's make sure minikube cluster is running fine and check if master node has taints applied on it.

kubectl describe  nodes minikube | grep -i taints



As you can currently no taints is applied on minikube master node , in this case any pod can be schedule on master node.  But in order to demonstrate , we will go ahead any apply taints on node.  We want to make sure that students with department=computer pod should only be schedule on node.

kubectl taint node minikube department=computer:NoSchedule



Taint is applied on master node and can be checked with describe command. Now Let's create a pod without toleration and deploy in cluster.

kubectl run student --image=nginx --dry-run=client --restart=Never  -o yaml > student.yaml

kubectl apply -f student.yaml

kubectl get pods


You might notice that student pod status goes into Pending state and it does not change as there is no node available for pod to schedule as master has started to repel pod.

Now let's apply toleration in student pod and try to schedule again. Edit student.yaml and add below highlighted section in yaml.


Save student.yaml definition file and re-create pod on minikube cluster.

kubectl delete pod student 
kubectl apply -f student.yaml



Once you do get pods , you can observe that pod is in running state and successfully scheduled on master since there was match of taints and toleration. 

Hope you find blog usefully. Keep learning and keep sharing..!!

Git Link : https://github.com/shashivish/kubernetes-example/tree/master/taintsAndtolerations

Reference : https://kubernetes.io/docs/home/ 






Comments

Post a Comment

Popular posts from this blog

JDBC Hive Connection fails : Unable to read HiveServer2 uri from ZooKeeper

-
Image
I have been struggling from couple of days to get my jdbc hive connection working using Zookeeper URI and after putting lot of effort finally I realised problem. In this post I will sharing my experience  and steps I took to rectify problem. Notes : I am having a HDP 2.6 sandbox along with my JDBC program . Use case : Many companies does not provide HiveServer2 url for configuring your JDBC string and its important because if your hiveserver goes down  , your whole job will abort (You will be missing Hive HA capabilities). So How do I dynamically determine second Hive Server URI ? Solution is to use Zookeeper Service Discovery feature. Zookeepers keeps all HiveServer 2 URI into its namespace. When client tries to make a connection to Hive , Zookeeper will look for available hive server and return URL to client program. You can quickly verify HiveServer namespace in zk Path with below command on HDP.  /usr/hdp/current/zookeeper-client/bin/zookeeper-client...
Read more

Access Kubernetes ConfigMap in Spring Boot Application

-
Image
 Hello There, Welcome back to our new post on Kubernetes tutorials. In this example we will see how can we configure and access ConfigMap in Spring boot application. ConfigMaps are helpful where you have multiple environment (Dev,Test Prod etc) and you want to promote your container with different environment configuration. There are three ways Pods can access ConfigMap in containers. For todays Demo , we will use Environment Variable approach. Pre-requisite : MiniKube Cluster   For Demo Purpose I have created Spring Boot application which will expose below end point.  curl localhost:8080/configMap Docker Image : shashivish123/springboot  Below is how Controller class looks like and value is being autowired from application.properties. Remember CONFIG_KEY value will be coming from Kubernetes ConfigMap Environment Configuration. Now Let's Begin , We will start by Creating firs ConfigMap in Kubernetes. Step 1: Create a namespace where we will ...
Read more

Developing Custom Processor in Apache Nifi

-
Image
Apache Nifi was developed to automate the flow of data between different systems. Apache NiFi is based on technology previously called “Niagara Files” that was in development and used at scale within the NSA for the last 8 years and was made available to the Apache Software Foundation through the NSA Technology Transfer Program. Nifi is based on FlowFiles which are heart of it. A FlowFile is a data record, which consists of a pointer to its content (payload) and attributes to support the content, that is associated with one or more provenance events. The attributes are key/value pairs that act as the metadata for the FlowFile, such as the FlowFile filename. The content is the actual data or the payload of the file. Provenance is a record of what’s happened to the FlowFile. Each one of these parts has its own repository (repo) for storage. Each flowfile is processed by FlowFile processor . Processors have access to attributes of a given FlowFile and its content stream. Processo...
Read more