[Bug]: Dokploy Cloud application schedule runs manually but never executes automatically

To Reproduce

1. Use Dokploy Cloud v0.30.0 with an application deployed to an active remote server.
2. Open the application's Scheduled Tasks section.
3. Create and enable an application schedule with:
   • Cron: */5 * * * *
   • Timezone: UTC
   • Shell: sh
   • Command: bun run cron:stripe-topup-reconcile
4. Wait for several scheduled intervals.
5. Observe that no automatic execution or deployment-history entry is created.
6. Trigger the same schedule manually.
7. Observe that the manual execution completes successfully.
8. Disable and re-enable the schedule, then save it again.
9. Change the cron expression to * * * * * and wait for several minutes.
10. Observe that automatic execution still never occurs.

Other scheduled tasks attached to the same application and remote server execute automatically.

Current vs. Expected behavior

Current behavior

The schedule record is saved correctly and reports enabled: true. Its cron expression and application association are also correct.

Manual execution works and produces a successful deployment record in approximately five seconds. Automatic execution creates no deployment record at all. This also occurs with * * * * *, so it is not specific to the five-minute expression.

Expected behavior

An enabled application schedule should be registered with the Dokploy Cloud scheduler and execute according to its cron expression. Each execution should create a deployment-history entry.

Provide environment information

Operating System:
    Control plane: Dokploy Cloud
    Remote server OS: Ubuntu 24.04.3 LTS
    Arch: Kernel: 6.8.0-100-generic (x86_64)
  Dokploy version: v0.30.0
  VPS Provider: Hetzner
  Application/service: Bun/TypeScript backend running in a Docker container
  Schedule type: Application
  Deployment target: Remote server

Which area(s) are affected? (Select all that apply)

Application

Are you deploying the applications where Dokploy is installed or on a remote server?

Remote server

Additional context

Dokploy API inspection confirms that the schedule is persisted with:

• enabled: true
• scheduleType: application
• the expected application association
• the exact submitted cron expression
• UTC as the default timezone

The deployment history contains only manually triggered executions. Other schedules attached to the same application execute automatically, which rules out a stopped application, unavailable remote server, or globally broken scheduler.

A Sentry Cron monitor independently reports every expected execution as missed.

Possible source-level lead: in the current Dokploy source, the Cloud paths in schedule.create and schedule.update invoke the asynchronous schedule(...) registration function without awaiting it:

https://github.com/Dokploy/dokploy/blob/canary/apps/dokploy/server/api/routers/schedule.ts

That function performs the request to the Cloud jobs service:

https://github.com/Dokploy/dokploy/blob/canary/apps/dokploy/server/utils/backup.ts

If registration fails or does not finish, the schedule database update can still succeed without surfacing the scheduler-registration failure.

Will you send a PR to fix it?

No

Start in apps/dokploy/server/api/routers/schedule.ts by reading the Cloud paths in schedule.create and schedule.update, then trace schedule(...) into apps/dokploy/server/utils/backup.ts and the Cloud jobs service request. Reproduce the application schedule case and verify that registration failures are surfaced and that an enabled schedule creates deployment-history entries during automatic execution.