Images with similar names (starts_with) do not get updated?

[keel-hq]

Images with similar names (starts_with) do not get updated?

[rusenask]

Hey, could you share your deployment.yaml files?
From the logs it looks like the opposite is happening of what you have described:

1. Event: msg="provider.kubernetes: processing event" registry= repository=ustwo/shifts tag=latest

2. This is skipped parsed_image_name="index.docker.io/ustwo/shifts_api:latest" (expected)

3. This is updated msg="provider.kubernetes: delta changed applied: ustwo/shifts:latest" (expected)

[arbarlow]

shifts deployment..

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  annotations:
    deployment.kubernetes.io/revision: "26"
    keel.sh/update-image: ustwo/shifts:latest
  creationTimestamp: 2017-08-08T14:49:37Z
  generation: 29
  labels:
    app: shifts
    keel.sh/policy: force
  name: shifts
  namespace: default
  resourceVersion: "134423"
  selfLink: /apis/extensions/v1beta1/namespaces/default/deployments/shifts
  uid: cc973a21-7c48-11e7-b30a-0667c411fdb7
spec:
  progressDeadlineSeconds: 600
  replicas: 1
  revisionHistoryLimit: 2
  selector:
    matchLabels:
      app: shifts
  strategy:
    rollingUpdate:
      maxSurge: 25%
      maxUnavailable: 25%
    type: RollingUpdate
  template:
    metadata:
      creationTimestamp: null
      labels:
        app: shifts
    spec:
      containers:
      - image: ustwo/shifts:latest
        imagePullPolicy: Always
        name: shifts
        ports:
        - containerPort: 80
          protocol: TCP
        resources: {}
        terminationMessagePath: /dev/termination-log
        terminationMessagePolicy: File
      dnsPolicy: ClusterFirst
      imagePullSecrets:
      - name: myregistrykey
      restartPolicy: Always
      schedulerName: default-scheduler
      securityContext: {}
      terminationGracePeriodSeconds: 30
status:
  availableReplicas: 1
  conditions:
  - lastTransitionTime: 2017-08-09T14:01:14Z
    lastUpdateTime: 2017-08-09T14:01:14Z
    message: Deployment has minimum availability.
    reason: MinimumReplicasAvailable
    status: "True"
    type: Available
  - lastTransitionTime: 2017-08-09T13:59:20Z
    lastUpdateTime: 2017-08-09T14:18:26Z
    message: ReplicaSet "shifts-1114249520" has successfully progressed.
    reason: NewReplicaSetAvailable
    status: "True"
    type: Progressing
  observedGeneration: 29
  readyReplicas: 1
  replicas: 1
  updatedReplicas: 1

shifts-api deployment..

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  annotations:
    deployment.kubernetes.io/revision: "8"
    keel.sh/update-image: ustwo/shifts_api:latest
  creationTimestamp: 2017-08-08T10:48:44Z
  generation: 14
  labels:
    app: shifts-api
    keel.sh/policy: force
  name: shifts-api
  namespace: default
  resourceVersion: "136403"
  selfLink: /apis/extensions/v1beta1/namespaces/default/deployments/shifts-api
  uid: 26033971-7c27-11e7-b30a-0667c411fdb7
spec:
  progressDeadlineSeconds: 600
  replicas: 2
  revisionHistoryLimit: 2
  selector:
    matchLabels:
      app: shifts-api
  strategy:
    rollingUpdate:
      maxSurge: 25%
      maxUnavailable: 25%
    type: RollingUpdate
  template:
    metadata:
      creationTimestamp: null
      labels:
        app: shifts-api
    spec:
      containers:
      - env:
        - name: POSTGRES
          value: postgresql://xxxxx
        image: ustwo/shifts_api:latest
        imagePullPolicy: Always
        name: shifts-api
        ports:
        - containerPort: 8000
          protocol: TCP
        resources: {}
        terminationMessagePath: /dev/termination-log
        terminationMessagePolicy: File
      dnsPolicy: ClusterFirst
      imagePullSecrets:
      - name: myregistrykey
      restartPolicy: Always
      schedulerName: default-scheduler
      securityContext: {}
      terminationGracePeriodSeconds: 30
status:
  availableReplicas: 2
  conditions:
  - lastTransitionTime: 2017-08-08T17:16:07Z
    lastUpdateTime: 2017-08-08T17:16:07Z
    message: Deployment has minimum availability.
    reason: MinimumReplicasAvailable
    status: "True"
    type: Available
  - lastTransitionTime: 2017-08-09T14:42:38Z
    lastUpdateTime: 2017-08-09T14:42:40Z
    message: ReplicaSet "shifts-api-3321753349" has successfully progressed.
    reason: NewReplicaSetAvailable
    status: "True"
    type: Progressing
  observedGeneration: 14
  readyReplicas: 2
  replicas: 2
  updatedReplicas: 2

[rusenask]

increased wait period to 5 seconds from 2 but basically k8s doesn't provide a nice way to force pull. If we encounter this issue again - maybe it would make sense to terminate pods when force pull is required. Of course that wouldn't be very graceful :)