CRITICAL: Instance record orphaned when enqueue fails after DB create

[poyrazK]

Why is this an issue?

In internal/core/services/instance.go:295-317, instance is created in DB first, then task enqueue fails. No cleanup of the orphaned DB record.

What is causing it?

if err := s.repo.Create(ctx, inst); err != nil {
    return nil, err
}

if err := s.taskQueue.Enqueue(ctx, "provision_queue", job); err != nil {
    return nil, errors.Wrap(errors.Internal, "failed to enqueue", err)
    // DB record orphaned - instance exists with STARTING status but no job
}

How can it be solved?

Add rollback on enqueue failure:

if err := s.taskQueue.Enqueue(...); err != nil {
    s.repo.Delete(ctx, inst.ID)  // Rollback DB record
    return nil, err
}

Category

• Small
• Medium
• Large

Severity

• Low
• Medium
• High
• Critical