Skip to content

backlog(B-0764): CNCF ecosystem as force multipliers behind Zeta interfaces (KEDA, DAPR, OPA, OAM/KubeVela) #5050

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
AceHack merged 1 commit into from
May 26, 2026
Merged

backlog(B-0764): CNCF ecosystem as force multipliers behind Zeta interfaces (KEDA, DAPR, OPA, OAM/KubeVela) #5050

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions docs/BACKLOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -699,6 +699,7 @@ are closed (status: closed in frontmatter)._
- [ ] **[B-0759](backlog/P2/B-0759-cluster-install-ux-audit-against-first-time-cli-user-persona-easier-than-proxmox-3-node-production-ready-aaron-2026-05-25.md)** Cluster-install UX audit against first-time-CLI-user persona — "easier than Proxmox" bar + 3-node production-ready inflection
- [ ] **[B-0761](backlog/P2/B-0761-zeta-cluster-as-open-source-reference-architecture-for-ai-to-train-on-and-compete-on-arc-agi-style-benchmark-aaron-2026-05-25.md)** Zeta cluster as open-source reference architecture for AI to train on and compete on — ARC-AGI-style benchmark substrate
- [ ] **[B-0762](backlog/P2/B-0762-ai-auto-submit-back-telemetry-fixes-from-in-the-wild-installs-adoption-cost-to-zero-flywheel-aaron-2026-05-25.md)** AI auto-submit-back telemetry + fixes from in-the-wild installs — adoption-cost-to-zero flywheel
- [ ] **[B-0764](backlog/P2/B-0764-cncf-ecosystem-as-force-multipliers-behind-zeta-interfaces-keda-dapr-opa-oam-kubevela-plus-ace-and-ontology-negotiation-aaron-2026-05-25.md)** CNCF ecosystem as force multipliers behind Zeta interfaces — KEDA, DAPR, OPA, OAM/KubeVela + Ace + ontology negotiation

## P3 — convenience / deferred

Expand Down
250 changes: 250 additions & 0 deletions ...eda-dapr-opa-oam-kubevela-plus-ace-and-ontology-negotiation-aaron-2026-05-25.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,250 @@
---
id: B-0764
priority: P2
status: open
title: CNCF ecosystem as force multipliers behind Zeta interfaces — KEDA, DAPR, OPA, OAM/KubeVela + Ace + ontology negotiation
effort: L
ask: aaron 2026-05-25
created: 2026-05-25
last_updated: 2026-05-25
depends_on:
- B-0741
- B-0763
composes_with:
- B-0747
- B-0748
- B-0749
- B-0754
- B-0759
- B-0761
- B-0762
tags: [cluster, cncf, plugins, keda, dapr, opa, oam, kubevela, ace, force-multipliers]
---

## Problem

Aaron 2026-05-25 mid-iteration-2-wait, extending B-0763
(negotiation-high-seat via owned interfaces): *"we can use things
like KEDA, all the different DAPR ecosystem and OPA that Open
application or whatever it was called for the kube cncf project,
things like that, and all that plus ../scratch like old schoold
package management of mangers plus ontology negoation turns all
those standards into force multipliers."*

The CNCF ecosystem has shipped enormous substrate that operators
benefit from but vendor lock-in models have not yet exploited:

| Project | Stage | What it provides |
|---|---|---|
| **KEDA** | CNCF Graduated | Event-driven autoscaling (queue depth, metrics, schedules → pod count) |
| **DAPR** | CNCF Incubating | Distributed-app building blocks (state, pub/sub, service-invoke, bindings, secrets, actors) as sidecars + SDK |
| **OPA** | CNCF Graduated | Policy-as-code via Rego (admission control, authz, config validation) |
| **OAM** + **KubeVela** | CNCF Sandbox | Application-model / Component-Trait separation (already filed at B-0749) |
| **Crossplane** | CNCF Incubating | Cloud-resource provisioning via k8s CRDs (already filed at B-0748) |
| **kro** | CNCF Sandbox | ResourceGraphDefinition + CEL composition (already filed at B-0748) |
| **Cilium** | CNCF Graduated | eBPF networking + service mesh + observability |
| **ArgoCD / Flux** | CNCF Graduated | GitOps reconciliation (already in Zeta substrate per B-0747) |
| **Longhorn** | CNCF Incubating | Replicated block storage (Zeta default per current substrate) |
| **Rook + Ceph** | CNCF Graduated | Storage orchestrator + distributed object store (Zeta future) |
| **Knative** | CNCF Incubating | Serverless on k8s |
| **OpenTelemetry** | CNCF Graduated | Observability standard (logs/metrics/traces) |

Each is well-engineered, battle-tested, has ecosystem momentum.
**Adopting them as plugins behind Zeta's interfaces gives Zeta
their substrate for free** while preserving the negotiation-high-
seat property from B-0763.

Combined with:

- **Ace** (Aaron's existing package-manager substrate — old-school
PM-of-PMs, per B-0247 / B-0287 / B-0288 lineage + B-0741 +
related Ace work)
- **Ontology negotiation** (B-0741): cross-cluster, cross-fork,
cross-vendor namespace bridging

...the CNCF ecosystem becomes a **force multiplier** for Zeta
rather than competition. Every CNCF project that ships becomes
another "plugin Zeta can offer operators behind a stable
interface."

## Target

Wire each major CNCF project into Zeta as a plugin behind a
Zeta interface (per B-0763 contract), so operators get:

- The CNCF project's substrate for free (KEDA's autoscaling,
DAPR's distributed-app patterns, OPA's policy engine, etc.)
- Zeta's stable operator-facing interface (operator code doesn't
change if the CNCF project upgrades, or if a non-CNCF
alternative is swapped in)
- Composition via Ace + ontology negotiation (operators declare
"I want autoscaling + policy + pub/sub" in Zeta-shape; the
Ace + ontology layer resolves to KEDA + OPA + DAPR; future
alternatives swappable)

## Acceptance

- [ ] **KEDA** plugin behind `Zeta.Scaling.EventDriven` interface
(CRD wrapper: workload + scaler config → KEDA
ScaledObject + TriggerAuthentication)
- [ ] **DAPR** plugin per building-block behind matching Zeta
interfaces:
- `Zeta.State.Store` (DAPR state component)
- `Zeta.Messaging.PubSub` (DAPR pubsub component)
- `Zeta.Service.Invoke` (DAPR service-to-service)
- `Zeta.Bindings.Input` / `Zeta.Bindings.Output`
- `Zeta.Secrets` (DAPR secret store component)
- `Zeta.Actors` (DAPR actor runtime)
- [ ] **OPA** plugin behind `Zeta.Policy.Engine` interface (Rego
policy evaluation; admission control via OPA Gatekeeper
OR built-in Kubernetes ValidatingAdmissionPolicy where
Rego is overkill)
- [ ] **OAM + KubeVela** plugin behind `Zeta.Application.Model`
interface (Component + Trait → KubeVela Application CRD);
composes with B-0749
- [ ] **Cilium** plugin behind `Zeta.Network.Mesh` interface
- [ ] **Knative** plugin behind `Zeta.Compute.Function`
interface (composes with B-0763 cloud-Function adapters
for serverless on k8s as the local-cluster option)
- [ ] **OpenTelemetry** plugin behind
`Zeta.Observability.{Metrics,Logs,Traces}` interfaces
(B-0763)
- [ ] **Rook + Ceph** plugin behind `Zeta.Storage.BlobStore` +
`Zeta.Storage.Block` interfaces (alternative to
Longhorn; operator swaps at cluster-build time or via
data migration)
- [ ] Ace + ontology negotiation integration: operator
declares desired capabilities in Zeta-shape; Ace
resolves to CNCF project install + config; ontology
bridge handles cross-vocabulary translation
- [ ] Documentation: `docs/cncf-ecosystem-as-plugins.md`
explaining the force-multiplier framing + per-project
plugin docs + swap paths
- [ ] Reference deployment: full Zeta cluster using KEDA +
DAPR + OPA + OAM + Knative + OpenTelemetry behind the
Zeta interfaces; documented + cost-estimated + working

## The force-multiplier framing

The substrate-honest argument: **adopting an ecosystem project
behind your own interface = you get their substrate + ecosystem
momentum + maintenance burden distribution, while keeping your
operator-facing contract stable**. Every CNCF project that
graduates adds force to your platform; you don't bear the cost of
building those primitives yourself.

The pattern composes with B-0763 negotiation-high-seat:

| Layer | What Zeta owns | What CNCF/vendors compete on |
|---|---|---|
| Operator API | `Zeta.<Capability>` interfaces | (Zeta-owned; stable contract) |
| Implementation | (chosen at install/runtime) | KEDA vs alternatives; DAPR vs alternatives; OPA vs alternatives; etc. |
| Underlying compute | (chosen at hardware/cloud) | NVMe vendors; cloud GPUs; etc. |

The operator sees only the Zeta interface; the CNCF project
ships substrate behind it; the vendor (cloud or on-prem)
ships hardware behind that. Three layers of competition; one
stable contract.

## Ace + ontology negotiation composition

Per Aaron's "package management of managers" framing:

- **Ace as PM-of-PMs**: the Ace package manager (per Aaron's
existing B-0247 / B-0287 / B-0288 / B-0741 substrate) is
designed to compose other package managers — npm, pip, gem,
cargo, helm, krew, etc. → Ace.
- **Ontology negotiation**: cross-vocabulary translation
between ecosystems (helm-charts ↔ kubevela-Components ↔
crossplane-Compositions ↔ kro-ResourceGraphDefinitions).
- **Together**: operator declares "I want a state store"; Ace
finds installed providers (DAPR / KEDA-managed Redis /
Cloudflare-KV via B-0763 plugin / etc.); ontology layer
translates between Zeta's `Zeta.State.Store` interface and
whichever provider's native API; operator code doesn't
change.

## Composes with

- B-0741 — ontology+category negotiation (the cross-vocabulary
bridge layer)
- B-0747 — git-native per-machine state + GitOps reconciliation
(the substrate the CNCF plugins reconcile against)
- B-0748 — kro/Crossplane/Koreo/middleware spectrum (the runtime
for declaring plugin choices via k8s CRDs)
- B-0749 — KubeVela/OAM Component/Trait (already filed; this row
references + composes)
- B-0754 — zero-typing first-boot (the install path needs CNCF
plugins discoverable at install time)
- B-0759 — first-time-CLI-user persona (CNCF plugin docs need
persona-aligned plain language; many CNCF docs are
expert-only)
- B-0761 — open reference architecture (the CNCF substrate
composition IS what makes the reference cloud-native; bare
k3s isn't a reference architecture, k3s + KEDA + DAPR + OPA
+ ArgoCD + OpenTelemetry IS)
- B-0762 — AI auto-submit-back telemetry (which CNCF plugin
combinations work best for which workloads — telemetry feeds
recommendations)
- B-0763 — cloud-native plugins fit Zeta interfaces (this row
is the CNCF-specific implementation of B-0763's general
pattern)
- Ace existing substrate (B-0247, B-0287, B-0288 + related) —
the PM-of-PMs that makes the plugin layer composable

## What this prevents

Without this scope, the failure mode is:

- Zeta builds its own autoscaling → competes with KEDA, loses
on ecosystem momentum
- Zeta builds its own pub/sub → competes with DAPR, loses on
feature parity
- Zeta builds its own policy engine → competes with OPA, loses
on Rego ecosystem
- Etc.

Each "build-it-yourself" decision burns engineering time +
loses the CNCF ecosystem's network effects. Adopting them as
plugins behind Zeta interfaces gets the ecosystem for free + lets
Zeta focus on the layer it uniquely owns: the operator-facing
interfaces + the install/upgrade/repair flow + the AI-native
substrate.

## Out of scope

- Building Zeta-native alternatives to any CNCF project — the
whole point is to NOT do that; adopt + integrate
- Tracking CNCF project lifecycle (graduated/incubating/sandbox
status changes) at row scope — handle via `docs/TECH-RADAR.md`
ring discipline
- Force-marketing Zeta to CNCF as a "we should be CNCF too"
pitch — premature; ship working substrate, let CNCF
recognition happen if it happens

## Strategic context

This row + B-0763 (cloud-native plugins) + B-0761 (open reference
architecture) + B-0762 (telemetry flywheel) + B-0741 (ontology
negotiation) compose into Zeta's full strategic substrate:

- **Own interfaces** (B-0763) → negotiation high seat
- **Adopt ecosystem** (B-0764) → force-multiplier behind those
interfaces
- **Open reference** (B-0761) → AI-trainable + competitively
benchmarked
- **Telemetry flywheel** (B-0762) → adoption-cost-to-zero
- **Ontology negotiation** (B-0741) → cross-vocabulary bridge

The competitive moat = the COMBINATION. Any one of these is
mimicable; the full stack composed coherently is not.

## Origin

Aaron 2026-05-25, mid-iteration-2 wait, extending B-0763's
negotiation-high-seat framing with the CNCF-ecosystem-as-force-
multiplier pattern. The OPA naming caught — Aaron initially said
"OPA that Open application or whatever it was called" which
conflates OPA (Open Policy Agent, Rego policy) with OAM (Open
Application Model, KubeVela). Both are referenced in this row;
both compose with the force-multiplier framing.
Loading