Velero: Creating or updating a schedule immediately starts a backup.

[Yggdrasil]

What steps did you take and what happened:

I deployed the below nightly backup schedule YAML via kubectl apply. This immediately triggered a new backup to start, instead of waiting for midnight! It appears the backup is also started at the scheduled time.

apiVersion: velero.io/v1
kind: Schedule
metadata:
  name: daily
  namespace: velero
spec:
  schedule: '@midnight'
  template:
    hooks: {}
    includedNamespaces:
    - '*'
    ttl: 720h0m0s

The same thing happens when an existing schedule is updated, for example when changing the schedule from @every 24h to @midnight. Instead of waiting till midnight, it starts right away then also runs at midnight.

What did you expect to happen:
I expected that creating the schedule resource do nothing, except create a cronjob. The schedule should then trigger only at the scheduled times. I.e. in case of @midnight it should start its initial run only when the time first passes midnight.

The output of the following commands will help us better understand what's going on:
Relevant lines are from kubectl logs deployment/velero -n velero. I edited the daily schedule to run at @midnight at 7:55 UTC. Noteworthy is the line: "Schedule is due, submitting Backup"

time="2020-04-14T06:11:20Z" level=info msg="Done checking for expired DeleteBackupRequests" controller=backup-deletion logSource="pkg/controller/backup_deleti
on_controller.go:484"
time="2020-04-14T07:11:20Z" level=info msg="Checking for expired DeleteBackupRequests" controller=backup-deletion logSource="pkg/controller/backup_deletion_co
ntroller.go:456"
time="2020-04-14T07:11:20Z" level=info msg="Done checking for expired DeleteBackupRequests" controller=backup-deletion logSource="pkg/controller/backup_deleti
on_controller.go:484"
time="2020-04-14T07:55:20Z" level=info msg="Schedule is due, submitting Backup" controller=schedule logSource="pkg/controller/schedule_controller.go:260" next
RunTime="2020-04-14 00:00:00 +0000 UTC" schedule=velero/daily
time="2020-04-14T07:55:20Z" level=info msg="Setting up backup log" backup=velero/daily-20200414075520 controller=backup logSource="pkg/controller/backup_contr
oller.go:472"
time="2020-04-14T07:55:20Z" level=info msg="Setting up backup temp file" backup=velero/daily-20200414075520 logSource="pkg/controller/backup_controller.go:494
"
time="2020-04-14T07:55:20Z" level=info msg="Setting up plugin manager" backup=velero/daily-20200414075520 logSource="pkg/controller/backup_controller.go:501"
time="2020-04-14T07:55:20Z" level=info msg="Getting backup item actions" backup=velero/daily-20200414075520 logSource="pkg/controller/backup_controller.go:505
"
time="2020-04-14T07:55:20Z" level=info msg="Setting up backup store" backup=velero/daily-20200414075520 logSource="pkg/controller/backup_controller.go:511"
time="2020-04-14T07:55:21Z" level=info msg="Writing backup version file" backup=velero/daily-20200414075520 logSource="pkg/backup/backup.go:213"
time="2020-04-14T07:55:21Z" level=info msg="Including namespaces: *" backup=velero/daily-20200414075520 logSource="pkg/backup/backup.go:219"
time="2020-04-14T07:55:21Z" level=info msg="Excluding namespaces: <none>" backup=velero/daily-20200414075520 logSource="pkg/backup/backup.go:220"
time="2020-04-14T07:55:21Z" level=info msg="Including resources: *" backup=velero/daily-20200414075520 logSource="pkg/backup/backup.go:223"
time="2020-04-14T07:55:21Z" level=info msg="Excluding resources: <none>" backup=velero/daily-20200414075520 logSource="pkg/backup/backup.go:224"

Environment:

• Velero version (use velero version):

Client:
    Version: v1.3.1
    Git commit: -
Server:
    Version: v1.3.1

• Velero features (use velero client config get features):
  features: <NOT SET>
• Kubernetes version (use kubectl version):

Client Version: version.Info{Major:"1", Minor:"17", GitVersion:"v1.17.3", GitCommit:"06ad960bfd03b39c8310aaf92d1e7c12ce618213", GitTreeState:"clean", BuildDate:"2020-02-13T18:06:54Z", GoVersion:"go1.13.8", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"14+", GitVersion:"v1.14.10-gke.27", GitCommit:"145f9e21a4515947d6fb10819e5a336aff1b6959", GitTreeState:"clean", BuildDate:"2020-02-21T18:01:40Z", GoVersion:"go1.12.12b4", Compiler:"gc", Platform:"linux/amd64"}

• Kubernetes installer & version:
  GKE 1.14.x
• Cloud provider or hardware configuration:
  Google Cloud Platform GKE

[skriss]

definitely agree this is a bug for schedules that are supposed to run at a fixed time each day, like the midnight schedule

[skriss]

Fixing this could be an opportunity to get rid of our own cron logic, and change the schedule controller to just create a Kubernetes CronJob.

[skriss]

this is actually a dupe of #1980, so I'm going to close this one out in favor of that existing issue.