deployment.yaml

#
#  Author: Hari Sekhon
#  Date: [% DATE  # 2019-11-25 13:47:16 +0000 (Mon, 25 Nov 2019) %]
#
#  vim:ts=2:sts=2:sw=2:et
#  lint: k8s
#
#  https://github.com/HariSekhon/Kubernetes-configs
#
#  License: see accompanying Hari Sekhon LICENSE file
#
#  If you're using my code you're welcome to connect with me on LinkedIn
#  and optionally send me feedback to help improve or steer this or other code I publish
#
#  https://www.linkedin.com/in/HariSekhon
#

# ============================================================================ #
#                              D e p l o y m e n t
# ============================================================================ #

# https://kubernetes.io/docs/concepts/workloads/controllers/deployment
#
# https://kubernetes.io/docs/reference/kubernetes-api/workload-resources/deployment-v1/

# This is your main point of deploying apps, and has lots of tricks and best practices baked in
#
# Edit / Delete as necessary for your deployment (it's easier than remembering / finding / adding all this stuff)

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: NAME
  namespace: NAMESPACE
  labels:
    app: APP
  annotations:
    # see also k8s_pod_disruption_budget.yaml
    cluster-autoscaler.kubernetes.io/safe-to-evict: "true"  # set to "false" for apps that don't like disruption, but that might prevent pods with emptyDir volumes getting rescheduled, preventing cluster autoscaler from scaling down the node pool
    # adds ArgoCD link symbol in the UI next to object
    #link.argocd.argoproj.io/external-link: http://www.domain.com/path
spec:
  paused: false
  minReadySeconds: 0  # secs healthy before serving
  progressDeadlineSeconds: 120
  revisionHistoryLimit: 10  # how many ReplicaSets specs to retain for rollbacks
  strategy:
    type: RollingUpdate
    rollingUpdate:
      # XXX: Do not set maxUnavailable to integers, you may cause outages in environments where the replicas <= maxUnavailable due to Dev environments with less replicas or where an HPA has scaled down
      maxUnavailable: 25%  # default: 25% - rounds down to int replicas
      maxSurge: 25%        # default: 25% - rounds up to int replicas - set higher for faster deployments for apps with long load times to ready eg. webapps with pre-loaded local cache
  selector:
    matchLabels:
      app: APP
      role: master
      tier: backend
  replicas: 3  # XXX: don't define replicas if using HPA, otherwise gets overridden and reset to this number every deploy which could cause load issues
  template:
    metadata:
      labels:
        app: APP
        role: master
        tier: backend
    spec:
      nodeSelector:
        disktype: ssd
        kubernetes.io/arch: amd64
        kubernetes.io/os: linux
        node.kubernetes.io/instance-type: m3.medium    # 1.17+
        # for persistent disks limited to 2 zones combined with GKE regional clusters
        topology.kubernetes.io/zone: europe-west2-a  # 1.17+
        topology.kubernetes.io/region: europe-west2    # 1.17+
        #myLabel: myFastNodes
        # GKE specific - hard requirement - prefer affinity below as safer soft requirement to still allow scheduling if node pool becomes unavailable (eg. recreated by Terraform)
        #cloud.google.com/gke-nodepool: POOL_NAME
      #priorityClassName: high-priority  # for business critical workloads to get priority access to the stable node pool instead of preemptible node pool, will evict lower priority pods to preemptible node pool (default 0) if necessary. Requires: priorityclass.yaml
      schedulerName: default-scheduler
      # https://kubernetes.io/docs/concepts/configuration/assign-pod-node/#affinity-and-anti-affinity
      # XXX: not recommended in clusters greater than several hundred nodes due to significant CPU overhead slowing down scheduling
      affinity:
        nodeAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
            - weight: 1  # 1 to 100, higher is stronger
              preference:
                matchExpressions:
                  - key: accelerator-type
                    operator: In
                    values:
                      - gpu
                      - tpu
              # XXX: prefer non-preemptible VMs unless unavailable
            - weight: 100
              preference:
                matchExpressions:
                  - key: cloud.google.com/gke-preemptible
                    operator: DoesNotExist
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
              - matchExpressions:
                  # AND's everything under this matchExpression
                  #
                  # XXX: remove this if the pod is safe to preempt
                  #
                  # Node preemption:
                  # - shuts down pods ungracefully
                  # - ignores PodDisruptionBudget
                  # - invalidates Stateful guarantees and can lead to data loss !!
                  # https://cloud.google.com/kubernetes-engine/docs/how-to/preemptible-vms#kubernetes_constraint_violations
                  #
                  # GCP: https://cloud.google.com/kubernetes-engine/docs/how-to/preemptible-vms#best_practices
                  - key: cloud.google.com/gke-preemptible
                    operator: DoesNotExist
                  # AWS: https://docs.aws.amazon.com/eks/latest/userguide/managed-node-groups.html
                  - key: eks.amazonaws.com/capacityType
                    operator: NotIn
                    values:
                      - SPOT
                  # Azure: https://docs.microsoft.com/en-us/azure/aks/spot-node-pool
                  - key: kubernetes.azure.com/scalesetpriority
                    operator: NotIn
                    values:
                      - spot
                  #
                  - key: kubernetes.io/e2e-az-name
                    operator: In # NotIn, Exists, DoesNotExist
                    # OR's the values
                    values:
                      - europe-west-1b
                      - europe-west-1c
        #podAffinity:
        podAntiAffinity:
          # XXX: spread app across AZs
          # XXX: use preferred if num_replicas > num_AZs
#          preferredDuringSchedulingIgnoredDuringExecution:
#          - weight: 100
#            podAffinityTerm:
#              topologyKey: topology.kubernetes.io/zone
#              labelSelector:
#                matchExpressions:
#                - key: app
#                  operator: In
#                  values:
#                  - APP
          requiredDuringSchedulingIgnoredDuringExecution:
              # XXX: spread app across AZs
            - topologyKey: topology.kubernetes.io/zone
              labelSelector:
                matchExpressions:
                  - key: app
                    operator: In
                    values:
                      - APP
              # XXX: spread app across hosts (default prefers spread, whereas this requires it)
            - topologyKey: kubernetes.io/hostname
              labelSelector:
                matchExpressions:
                  - key: app
                    operator: In
                    values:
                      - APP
      # ignore taints
      #tolerations:
        #- key: "myKey"
        #  operator: "Exists"  # if myKey exists ignore NoSchedule
        #  operator: "Equal"   # if myKey's value = "value" ignore NoSchedule
        #    value: "value"
        #  effect: "NoSchedule"
      # XXX: allow scheduling on preemptible VMs with a NoSchedule taint if safe to do so
      tolerations:
        # XXX: this must be set yourself on AWS EKS / GCP GKE
        # taint AWS spot / GCP preemptible nodepools with short 'preemptible=true:NoSchedule' taint for convenience
        #- key: cloud.google.com/gke-preemptible
        - key: preemptible
          operator: Equal
          value: "true"
          effect: NoSchedule
          # Azure auto-adds this taint, no need to create manually
        - key: kubernetes.azure.com/scalesetpriority
          operator: Equal
          value: spot
          effect: NoSchedule
      terminationGracePeriodSeconds: 120  # how long between SIGTERM and SIGKILL (kill -9) - may want to set higher than default 30 secs eg. 1800 for Cassandra
      dnsPolicy: ClusterFirst # use coredns, Default uses node's dns resolution
      restartPolicy: Always
      #serviceAccountName: NAME
      securityContext:
        # see pod-security-policy.yaml for wider enforcement
        runAsNonRoot: true
        #runAsUser: 0   # not recommended
        fsGroup: 1000   # group to own volumeMounts so they can be written to by the user-level process
        # XXX: don't chgrp all the many small files unless the root mismatches, prevents long outages on pod creation hanging with multiple of these error messages:
        #
        #   Unable to attach or mount volumes: unmounted volumes=[jenkins-home], unattached volumes=[kube-api-access-6q2kb jenkins-home]: timed out waiting for the condition
        #
        #   https://docs.cloudbees.com/docs/cloudbees-ci-kb/latest/cloudbees-ci-on-modern-cloud-platforms/timeout-when-attaching-volumes-in-kubernetes
        #
        fsGroupChangePolicy: OnRootMismatch
      initContainers:
        - name: init-files
          image: alpine/git:latest
          command: ['git', 'clone', 'https://github.com/HariSekhon/DevOps-Python-tools /data']
          volumeMounts:
            - name: app-vol
              mountPath: /data
        - name: init-mysql-service
          image: busybox:latest
          command: ['sh', '-c', 'until nslookup mysql; do echo waiting for mysql service DNS to come up; sleep 1; done']
        - name: init-mysql-endpoint
          image: busybox:latest
          command: ['sh', '-c', 'until nc -z mysql 3306; do echo waiting for mysql endpoint port to come up; sleep 1; done']
      containers:
        - name: master
          image: k8s.gcr.io/APP:latest
          imagePullPolicy: IfNotPresent  # set to Always if replacing a docker_image:tag or using :latest in development
          ports:
            - name: http
              containerPort: 8080
              protocol: TCP
          env:
            - name: HEAP_SIZE
              value: 1G
            - name: CASSANDRA_SEEDS
              value: cassandra-0.cassandra.svc.cluster.local,cassandra-1.cassandra.svc.cluster.local,cassandra-2.cassandra.svc.cluster.local
            - name: POD_IP
              valueFrom:
                fieldRef:
                  fieldPath: status.podIP  # K8S API path
            - name: CLOUD_SQL_INSTANCE
              valueFrom:
                configMapKeyRef:
                  key: cloud-sql-instance
                  name: my-configmap
            - name: MYSECRET
              valueFrom:
                secretKeyRef:
                  name: my-secret
                  key: my-secret
          readinessProbe:
            # http, tcp or exec
            httpGet:
              path: /healthz
              port: 8080
              #scheme: HTTPS
              #insecureSkipTLSVerify: true
            # only uncomment if you can't use http check, must delete http check otherwise will get this error:
            # invalid: spec.containers[0].livenessProbe.tcpSocket: Forbidden: may not specify more than 1 handler type
            #tcpSocket:
            #  port: 8080
            #exec:
            #  command:
            #    - pgrep -f somecommand
            initialDelaySeconds: 0 # default
            successThreshold: 1    # default
            failureThreshold: 3    # default
            periodSeconds: 10      # default (interval)
            timeoutSeconds: 1      # default
          livenessProbe:
            httpGet:
              path: /healthz
              port: 8080
              #scheme: HTTPS
              #insecureSkipTLSVerify: true
            #tcpSocket:
            #  port: 8080
            #exec:
            #  command:
            #    - pgrep -f somecommand
            initialDelaySeconds: 0 # default
            successThreshold: 1    # default
            failureThreshold: 3    # default
            periodSeconds: 10      # default (interval)
            timeoutSeconds: 1      # default
          resources:
            requests:
              # millicores - 1/1000 of a CPU / vCPU
              # 100m = 1/10th of CPU core
              cpu: 100m
              memory: 100Mi # MiB
            limits:
              cpu: "1"
              memory: 4Gi  # GiB
          volumeMounts:
            - name: APP-vol
              mountPath: /mnt/APP-vol
              readOnly: true
            - name: env
              mountPath: /app/.env
              subPath: .env
              readOnly: true
          securityContext:
            runAsNonRoot: true
            privileged: false
            allowPrivilegeEscalation: false
            readOnlyRootFilesystem: true
            capabilities:
              drop:
                - ALL
              add: ["IPC_LOCK"]  # allow mlock to avoid paging ram to disk, for security or performance eg. NoSQL low latency datastores
          lifecycle:
           #postStart:
           #  exec:
           #    command: ["/bin/sh", "-c", "echo Hello from the postStart handler > /usr/share/message"]
            # execute this when sending TERM to container
            preStop:
              # http call or exec
              exec:
                # recent nginx-ingress already has this, but still a decent example to gracefully finish connections on your own nginx via SIGQUIT not SIGTERM
                # also remember to increase terminationGracePeriodSeconds
                #command: ["/bin/sh","-c","nginx -s quit; while killall -0 nginx; do sleep 1; done"]
                command:
                  - /bin/sh
                  - -c
                  - nodetool drain
          terminationMessagePath: /dev/termination-log
          terminationMessagePolicy: File  # FallbackToLogsOnError to use container logs if message file is empty upon container error
        #
        # GCP Cloud SQL Proxy sidecar - app can then connect to DB at localhosta (see also cloud-sql-proxy*.yaml for a standalone shared deployment & service)
        #
        #   https://cloud.google.com/sql/docs/postgres/connect-kubernetes-engine
        #
        # Set up Workload Identity integration between k8s SA + GCP SA
        #
        #   https://cloud.google.com/sql/docs/postgres/connect-kubernetes-engine#workload-identity
        #
        - name: cloud-sql-proxy
          image: gcr.io/cloudsql-docker/gce-proxy:1.17
          command:
            - /cloud_sql_proxy
            - "-ip_address_types=PRIVATE"
            # XXX: trade off exposing listener on 0.0.0.0 required for having a kubernetes health check
            - "-instances=$(CLOUD_SQL_INSTANCE)=tcp:5432"  # populate CLOUD_SQL_INSTANCE env var via configmap, must also specify a port (eg. 5432 for PostgreSQL or 3306 for MySQL, or else a -dir option if using socket dir)
          securityContext:
            # The default Cloud SQL proxy image runs as the
            # "nonroot" user and group (uid: 65532) by default.
            runAsNonRoot: true
          env:
            - name: CLOUD_SQL_INSTANCE
              valueFrom:
                configMapKeyRef:
                  # XXX: Edit
                  name: my-configmap
                  key: cloud-sql-instance
          readinessProbe:
            tcpSocket:
              #port: 3306  # mysql
              port: 5432   # postgres
            periodSeconds: 5
          livenessProbe:
            tcpSocket:
              #port: 3306  # mysql
              port: 5432   # postgres
            failureThreshold: 10
          resources:
            requests:
              cpu: 100m
              memory: 128mi
            limits:
              cpu: "1"
              memory: 2gi
      volumes:
        - name: APP-vol
          # GKE will automatically create the underlying disk for us using default storage provider's dynamic provisioner
          persistentVolumeClaim:
            claimName: APP-dynamic-pvc
        - name: env
          configMap:
            name: myconfigmap
            defaultMode: 0400
            items:
              - key: .env
                path: .env