Running applications in Pods
spring petclinic
- This application is developed in java using spring boot
- this application runs on port 8080
- image:
shaikkhajaibrahim/spcjan2024:1.0 - port: 8080
- docker command
docker run --name spc -d -P shaikkhajaibrahim/spcjan2024:1.0
* Refer Here for pod specification with port info added
* Note: Exposing Pod directly to external world is not a recommended practice, we are using the following only for evaluation
* Generally while creating containers it is a good idea to define the cpu and memory limits Refer Here for article
docker run --name spc -d -P --memory "512m" --cpus="0.5" shaikkhajaibrahim/spcjan2024:1.0
- Kubernetes also has limits and requests to support allocating upper and lower bound for memory, cpu and other restrictions Refer Here
- Refer Here for the changes done in the specification to include
- lower limits for cpu and memory
- upper limits for cpu and memory
- Label selector Refer Here
-
Lets create multiple pods i.e. lets use replicaset Refer Here for changes done
Service
- Service is a k8s object which creates a virtual ip address and a DNS Name which is fixed.
- To make this service accessible kube-proxy and CORE DNS will do the job of forwarding/making service accessible
- in every container we have a file
/etc/resolve.conf - Services are of different types
- Cluster Ip (ip address within k8s cluster)
- Node Port (Expose service on some port on all nodes of k8s cluster)
- Load Balancer (Used in Managed k8s clusters)
- ExternalName (It gives a DNS Record)
- when we create a service, it creates endpoints, Each endpoint represents a podip with port
- Refer Here for service
-
Refer Here for the changes done
-
Overview
