Docker best practices practices

Within the framework of this article, the most frequent antipatterns in the design of docker images will be considered, and the optimal solution for each will be presented. Using a redundant base image One of the common antipatterns when using Docker is the use of a redundant base image. # Bad FROM ubuntu The optimal solution is to use the most lightweight base image, which contains only the necessary components, which can act as alpine....

October 8, 2023 · 3 min · OldTyT

Docker for the youngest

To consolidate the material, it is recommended to execute the commands specified in the article Introduction Docker is a platform for developing, delivering, and running applications in containers. Containers are lightweight and isolated environments that allow you to run applications on any operating system without the need for additional dependencies. In this article, we will explore the basics of Docker and demonstrate how to use it for application development. Installing Docker Before you can start using Docker, you need to install it on your system....

October 8, 2023 · 2 min · OldTyT

How I launched minecraft in kubernetes

Minecraft in kubernetes Hello everyone, in this article I will share with you my experience in launching minecraft in kubernetes. The main problem that arises when trying to launch minecraft in kubernetes is that while minecraft is running, files are being worked on on the local disk, and this approach in kubernetes is anti-pathetic, since in case of failure of the node (let’s call it node A) on which the pod was deployed - pod will not be able to be created on another node (node B), since it is known that pod previously worked on node A and, accordingly, all its permanent files are there. Options for solving problems with the use of permanent files While searching for a solution to this problem, I identified the following solutions for myself: rejection of permanent files use native Persistent Volume create a network drive use csi-s3 Let’s take a closer look at each of the options Abandoning permanent files Positive: The best solution for applications in kubernetes Minuses: Not applicable to the current application Use Native Persistent Volume Positive: High speed of reading/writing files Minuses: All data is on the same node, in case of its breakdown, the pod will not be able to rise Create a network drive Positive: The data is replicated to multiple nodes Minuses: Low file read/write speed Use csi-s3 Positive: The data was taken out from the cluster Minuses: Low file read/write speed An extremely high price may come out due to the large number of API calls Analysis of options Based on the data obtained, we come to the conclusion that at the moment, there is no solution that could meet our requirements, namely: high speed of reading / writing files there must be no binding to a specific node it should be done as budget-friendly as possible Solving the problem using persistent data To solve this problem, was prepared docker container. The algorithm of the container SSH keys are being added. The keys are taken from the values of the ENV variables. The repository is cloned from github with the server configuration to the directory - /app The sequences MYSQL_.+ are replaced in all files in /app/plugins by their value. The game world is being copied from s3 storage and unpacked Game plugins and kernel are being copied from s3 storage Starts /task_manager.py - which is responsible for the execution of cron tasks, the operation of the server and the output of syslog Cron tasks In cron tasks, the following tasks are currently configured: * * * * * root /git_pull.sh | logger * * * * * root /git_commit.sh | logger 0 * * * * root /copy_worlds.sh | logger Learn more about each task: /git_pull.sh - executes git pull in the /app directory /git_commit.sh - creates a commit and pushes it to the repository /copy_worlds.sh - makes a copy of the world and uploads it to s3 storage in place of the existing one Example of service deployment Setting up the ingress controller Consider the example of traffic'. To configure the ingress` controller, you will need to run the following command: $ helm upgrade --values traefik-values.yaml traefik traefik/traefik -n kube-system Contents of the traefik-values.yaml file: ports: traefik: port: 9000 expose: false exposedPort: 9000 protocol: TCP web: port: 8000 expose: true exposedPort: 80 protocol: TCP minecraft: port: 22565 expose: true exposedPort: 25565 protocol: TCP websecure: port: 8443 expose: true exposedPort: 443 protocol: TCP http3: enabled: false tls: enabled: true options: "" certResolver: "" domains: [] middlewares: [] metrics: port: 9100 expose: false exposedPort: 9100 protocol: TCP After that, connections to LoadBalancer will be allowed on port 25565: $ kubectl get svc -n kube-system | grep -v "none" NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE traefik LoadBalancer 10.43.2.50 100.10.0.11 25565:30224/TCP,80:31380/TCP,443:30226/TCP 2d21h Deploy Example of the final manifest that contains the required configuration: --- apiVersion: v1 kind: Namespace metadata: name: minecraft --- apiVersion: v1 kind: Secret metadata: name: minecraft-secret namespace: minecraft type: Opaque data: SSH_KEY_PUBLIC: "SECRET" SSH_KEY_PRIVATE: "SECRET" MYSQL_ROOT_PASSWORD: "SECRET" MYSQL_DBS: "SECRET" MYSQL_HOST: "SECRET" MYSQL_USER: "SECRET" MYSQL_PASSWORD: "SECRET" S3_BUCKET: "SECRET" S3_ACCESS_KEY: "SECRET" S3_ACCESS_KEY_ID: "SECRET" --- apiVersion: apps/v1 kind: Deployment metadata: name: some_worlds_name namespace: minecraft spec: selector: matchLabels: app: some_worlds_name strategy: type: Recreate template: metadata: namespace: minecraft labels: app: some_worlds_name spec: imagePullSecrets: - name: github-registry containers: - image: ghcr.io/oldtyt/docker_minecraft imagePullPolicy: Always name: some_worlds_name resources: requests: cpu: 1000m memory: 5G limits: memory: 6G cpu: 1200m env: - name: XMX value: "5G" - name: "XMS" value: "512M" - name: MYSQL_HOST valueFrom: secretKeyRef: name: minecraft-secret key: MYSQL_HOST - name: MYSQL_USER valueFrom: secretKeyRef: name: minecraft-secret key: MYSQL_USER - name: MYSQL_DB valueFrom: secretKeyRef: name: minecraft-secret key: MYSQL_USER - name: MYSQL_PASSWORD valueFrom: secretKeyRef: name: minecraft-secret key: MYSQL_PASSWORD - name: S3_BUCKET valueFrom: secretKeyRef: name: minecraft-secret key: S3_BUCKET - name: S3_ACCESS_KEY valueFrom: secretKeyRef: name: minecraft-secret key: S3_ACCESS_KEY - name: S3_ACCESS_KEY_ID valueFrom: secretKeyRef: name: minecraft-secret key: S3_ACCESS_KEY_ID - name: SSH_KEY_PRIVATE valueFrom: secretKeyRef: name: minecraft-secret key: SSH_KEY_PRIVATE - name: SSH_KEY_PUBLIC valueFrom: secretKeyRef: name: minecraft-secret key: SSH_KEY_PUBLIC - name: PLUGINS_LIST value: "some,plugins,list" - name: GIT_REPO value: "git@github.com:USER/REPO.git" - name: KERNEL value: "KERNEL" - name: "WORLDS" value: "some_worlds_name" --- apiVersion: v1 kind: Service metadata: name: some_worlds_name namespace: minecraft labels: env: prod app: some_worlds_name owner: OldTyT spec: ports: - name: minecraft targetPort: 25565 port: 25565 protocol: TCP selector: app: some_worlds_name The command to deploy the manifest: $ kubectl apply -f minecraft.yaml

June 5, 2023 · 5 min · OldTyT