kapp unblocks dependent changes if the changes which are depended upon already exists

[praveenrewar]

What steps did you take:
jobs.yaml

apiVersion: batch/v1
kind: Job
metadata:
  name: successful-job-1
  annotations:
    kapp.k14s.io/change-rule: "upsert after upserting abc"
spec:
  template:
    metadata:
      name: successful-job-1
    spec:
      containers:
        - name: successful-job-1
          image: busybox
          command: ["/bin/sh", "-c", "sleep 10"]
      restartPolicy: Never
---
apiVersion: batch/v1
kind: Job
metadata:
  name: successful-job-2
  annotations:
    kapp.k14s.io/change-group: "abc"
spec:
  template:
    metadata:
      name: successful-job-2
    spec:
      containers:
        - name: successful-job-2
          image: busybox
          command: ["/bin/sh", "-c", "sleep 10"]
      restartPolicy: Never

Deployed these jobs and stopped in between to create only job 1 without waiting for it.

kapp deploy -a order -f order.yaml
Target cluster 'https://192.168.64.12:8443' (nodes: minikube)

Changes

Namespace  Name              Kind  Age  Op      Op st.  Wait to    Rs  Ri
default    successful-job-1  Job   -    create  -       reconcile  -   -
^          successful-job-2  Job   -    create  -       reconcile  -   -

Op:      2 create, 0 delete, 0 update, 0 noop, 0 exists
Wait to: 2 reconcile, 0 delete, 0 noop

Continue? [yN]: y

6:28:20PM: ---- applying 1 changes [0/2 done] ----
6:28:20PM: create job/successful-job-2 (batch/v1) namespace: default
6:28:20PM: ---- waiting on 1 changes [0/2 done] ----
6:28:20PM: ongoing: reconcile job/successful-job-2 (batch/v1) namespace: default
6:28:20PM:  ^ Waiting to complete (0 active, 0 failed, 0 succeeded)
6:28:20PM:  L ongoing: waiting on pod/successful-job-2-jlwbz (v1) namespace: default
6:28:20PM:     ^ Pending: ContainerCreating
^C

Deploy the jobs again

kapp deploy -a order -f order.yaml
Target cluster 'https://192.168.64.12:8443' (nodes: minikube)

Changes

Namespace  Name              Kind  Age  Op      Op st.  Wait to    Rs       Ri
default    successful-job-1  Job   -    create  -       reconcile  -        -
^          successful-job-2  Job   5s   -       -       reconcile  ongoing  Waiting to complete (1 active, 0
                                                                            failed, 0 succeeded)

Op:      1 create, 0 delete, 0 update, 1 noop, 0 exists
Wait to: 2 reconcile, 0 delete, 0 noop

Continue? [yN]: y

6:28:27PM: ---- applying 2 changes [0/2 done] ----
6:28:27PM: noop job/successful-job-2 (batch/v1) namespace: default
6:28:27PM: create job/successful-job-1 (batch/v1) namespace: default
6:28:27PM: ---- waiting on 2 changes [0/2 done] ----
6:28:27PM: ongoing: reconcile job/successful-job-2 (batch/v1) namespace: default
... snip ...
6:28:48PM:  ^ Completed
6:28:48PM: ---- applying complete [2/2 done] ----
6:28:48PM: ---- waiting complete [2/2 done] ----

Succeeded

What happened:
Even though successful-job-2 depends on successful-job-1 kapp tried to create successful-job-2 before waiting for successful-job-1 if the operation is noop for it.

What did you expect:
kapp should have waited for successful-job-1 to be created successfully before deploying successful-job-2.

Environment:

• kapp version (use kapp --version):
• OS (e.g. from /etc/os-release):
• Kubernetes version (use kubectl version)

---

Vote on this request

This is an invitation to the community to vote on issues, to help us prioritize our backlog. Use the "smiley face" up to the right of this comment to vote.

👍 "I would like to see this addressed as soon as possible"
👎 "There are other more important things to focus on right now"

We are also happy to receive and review Pull Requests if you want to help working on this issue.

[github-actions]

This issue is being marked as stale due to a long period of inactivity and will be closed in 5 days if there is no response.