Interpretation (optional)
Customer: customer-easterwood
What's Fleet missing?
When creating a new team via GitOps that includes VPP apps, software, and other resources, the configuration cannot be applied atomically in a single fleetctl gitops run. Instead, users must run GitOps multiple times (e.g., once to create the team, again to add VPP apps, again for other items) because dependent resources fail when the team doesn't yet exist during the same run.
Current behavior:
- User defines a new team in their GitOps YAML with VPP apps, software, profiles, etc.
- Runs
fleetctl gitops
- The team is created, but VPP app assignments and some other resources fail because they reference a team that was just created in the same run
- User must re-run
fleetctl gitops 2-3 more times to get everything applied
- This is compounded by GitOps runs already taking 20-25 minutes for environments with 20+ teams
Ideal workflow:
- User declares the full desired state for a new team in YAML (team settings, VPP apps, software, profiles, scripts — everything)
- Runs
fleetctl gitops once
- Fleet handles the ordering internally (create team → assign resources) within a single run
- The team is fully configured after one pass
Why this matters:
Customer has 20+ teams organized by region and business entity. Each GitOps run already takes 20-25 minutes. Requiring 3 separate runs to onboard a new team means ~1 hour of pipeline time for what should be a single declarative operation. This undermines the "declare desired state" promise of GitOps.
Context from Gong (Jan 22, 2026):
Former CSA Jason Lewis noted he wanted to investigate whether this is a bug or intentional. Customer (Jarryd Stanbrook, Easygo) described wanting to "declare their ideal state, even if it means creating a team and adding all apps and VPP stuff in one go, rather than in three separate iterations."
Related issues:
customer-coelestina: https://fleetdm.slack.com/archives/C09B2NPQR8V/p1777064602519059customer-easterwood: Gong snippet: Recording Jan 22, 2026Interpretation (optional)
Customer: customer-easterwood
What's Fleet missing?
When creating a new team via GitOps that includes VPP apps, software, and other resources, the configuration cannot be applied atomically in a single
fleetctl gitopsrun. Instead, users must run GitOps multiple times (e.g., once to create the team, again to add VPP apps, again for other items) because dependent resources fail when the team doesn't yet exist during the same run.Current behavior:
fleetctl gitopsfleetctl gitops2-3 more times to get everything appliedIdeal workflow:
fleetctl gitopsonceWhy this matters:
Customer has 20+ teams organized by region and business entity. Each GitOps run already takes 20-25 minutes. Requiring 3 separate runs to onboard a new team means ~1 hour of pipeline time for what should be a single declarative operation. This undermines the "declare desired state" promise of GitOps.
Context from Gong (Jan 22, 2026):
Former CSA Jason Lewis noted he wanted to investigate whether this is a bug or intentional. Customer (Jarryd Stanbrook, Easygo) described wanting to "declare their ideal state, even if it means creating a team and adding all apps and VPP stuff in one go, rather than in three separate iterations."
Related issues: