Der Shop ist online

Wir bauen Workstation AI-Maschinen für Websites, KI-Agenten und LLM-Training – online konfigurieren und kaufen.

Händlerpreise erhalten
KI-Produkte
KI-LabsOpenAI-AgentenClaude-AgentenGrok-BotWorkstation AI-natives CRMMarketingAlle Produkte
KI-Lösungen
Angewandte KIKI-AgentenGenerative KIKI-AutomatisierungKI-WeiterbildungKI-Pakete für KMUEnterprise AI LabKI-WorkstationsNVIDIA-Hardware & -SoftwareNVIDIA Accelerated ComputingPrivate KIGPU-Private-CloudGPU-ClusterEdge-KIKI nach Branche
KI-Services
PlattformmodernisierungDigital EngineeringDatengrundlagen & KIAutonomer BetriebKI-BeratungDevOps-AutomatisierungCybersicherheitSoftwareentwicklungAgentenentwicklungMLOps-EinrichtungGPU-LeasingGPU-Private-Cloud
Über uns
UnternehmenUnsere GeschichtePartnerKundengeschichten
KI-Rechenzentren
KI-RechenzentrenGPU-Private-CloudGPU-ClusterGPU-Leasing
KI-EinblickeJev-ReiheBenchmark LabModellauswahl
Technische Hinweise
Dokumentation
Linux-BetriebssystemeUbuntuRed Hat Enterprise LinuxSUSE LinuxAlpine LinuxDebian, Rocky & AlmaLinux
KI-Cloud-PlattformenAWS & Amazon BedrockGoogle Cloud & Vertex AIMicrosoft Azure & AI FoundryGPU-Clouds & Marktplätze
Private KI-RechenzentrenFür AnwaltskanzleienFür Steuer- und WirtschaftsprüfungskanzleienOn-Premises-PlattformenVollständig abgesicherte Cloud
ToolsRing PromoterJobshout
News
Blog
KontaktAnmelden

KI-Workstations, Multi-Agenten-KI-Software, GPU-Infrastruktur und intelligente Agentenlösungen für moderne Unternehmen.

Kontakt

KI-Lösungen

Angewandte KIKI-AgentenGenerative KIKI-AutomatisierungKI-WeiterbildungKI-Pakete für KMUEnterprise AI LabKI nach Branche

Produkte

Alle ProdukteWorkstation AI-natives CRM & ERPMarketingOpenAI-AgentenRing PromoterJobshout

Unternehmen

Über unsUnternehmenKI-RechenzentrenWarum WorkstationPartnerKundengeschichtenPreiseKontaktTenthMatrix

Ressourcen

KI-EinblickeTechnische HinweiseDokumentationLinux-BetriebssystemeKI-Cloud-PlattformenPrivate KI-RechenzentrenBlogSucheSitemap
Büro Belgien
Workstation SRL, Rue Vanderkindere 34, 1180 Uccle, BrusselsBE 0751.518.683Mo – Fr: 9:00 – 18:00 Uhr MEZ
+32 492 45 67 46
Hauptsitz Großbritannien
77-79 Marlowes, Hemel Hempstead HP1 1LFAnfahrt – Ausfahrt 20 der M25 (Outer London)Handelsregisternr.: 11641870Mo – Fr: 9:00 – 18:00 Uhr GMT
+44 7515 356 146
Büro Indien
#159 Sector 9, Pocket 1, DDA Flats, 110077 Dwarka, New Delhi
+91 98881 98841
Büro Taiwan
3F, No. 2-4, Aisi Rd., Ren'ai Dist., Keelung City 200, Taiwan基隆市仁愛區愛四路2之4號3樓Mo – Fr: 9:00 – 18:00 Uhr Taiwan-Zeit (GMT+8)
+886 926 892 701+32 492 45 67 46
Büro Estland
Aleksander Puškini tn 30-25, 20306 Narva, EstoniaMo - Fr: 9:00 - 18:00 Uhr estnischer Zeit (EET)
+32 492 45 67 46

© 2026 Workstation AI. Alle Rechte vorbehalten.

DatenschutzDatenlöschungCookiesNutzungsbedingungenWebsite-Sitemap

NVIDIA, das NVIDIA-Logo, RTX, RTX PRO, Blackwell, CUDA, NVLink, MGX, Omniverse, Spectrum-X, Quantum, BlueField, DGX und HGX sind Marken und/oder eingetragene Marken der NVIDIA Corporation in den USA und anderen Ländern. AMD, das AMD-Arrow-Logo, EPYC, Instinct, Ryzen und Threadripper sind Marken von Advanced Micro Devices, Inc. Dell und PowerEdge sind Marken der Dell Inc. oder ihrer Tochtergesellschaften. Intel und Xeon sind Marken der Intel Corporation oder ihrer Tochtergesellschaften. AMAX, AceleMax und LiquidMax sind Marken der AMAX Engineering Corporation. Alle anderen Produktnamen, Logos und Marken sind Eigentum ihrer jeweiligen Inhaber. Wir verwenden sie ausschließlich, um die Produkte zu kennzeichnen, die wir beschreiben, liefern oder unterstützen; ihre Verwendung bedeutet nicht, dass ihre Inhaber uns empfehlen.

Startseite / Technische Hinweise / Technology
DevOpsCI/CDAIKubernetesAutomation

Ring Promoter: Modern CI/CD You Cannot Miss for AI-Powered Deployments

Technical brief: ring promotion control plane, version-verified health, kubectl / GitHub Actions / k8sjob deployers, and AI-powered deployment workflows

15. August 2026Technology5 min read

Watch: Ring Promoter quick intro (5 min)

Ring Promoter — AI-assisted promotion to production (5-minute intro)

5-minute intro: youtu.be/SjXwaTVuYhM · Product: ringpromoter.com

Ring Promoter is Workstation’s modern CI/CD promotion control plane — the tool you cannot miss for AI-powered deployments. It moves versions of many applications through ordered deployment rings int → test → acc → prod with live health gates, optional version verification, automatic rollback, and a full audit trail. Product site: https://www.ringpromoter.com/ · source: github.com/bwalia/ring-promoter. Companion: blog · product page: /ring-promoter.

Ring Promoter cover — modern CI/CD for AI deployments

Agent digest.
  • What: Shared ring pipeline for multi-app promotion with health-gated hops and auto-rollback.
  • Who: Platform, SRE, DevOps, and AI/ML teams shipping agents, APIs, gateways, and classic services.
  • Control: Embedded web UI + JSON REST API; Postgres history per (app, ring).
  • Deployers: kubectl, GitHub Actions workflow-dispatch, or Kubernetes Job runners — per app.
  • Tagline: Every release earns production.

Watch: Ring Promoter demo

Ring Promoter — Every release earns production (YouTube demo)

Workstation demo: youtu.be/IzhvmrwEebE · Product: ringpromoter.com

1. Positioning: CI/CD for the AI shipping era

AI platforms do not ship like a single monolith. You promote inference services, agent runtimes, MCP gateways, RAG APIs, and supporting microservices on different clocks — often from different repos and different deploy mechanisms (Kubernetes vs VM CI). Classic “merge to main and hope” pipelines skip rings, fake green health, and leave no shared promotion protocol across apps.

Ring Promoter is the missing control plane: one ordered set of rings, many applications, strict promotion rules, and swappable deployers. It is small enough to run on k3s, serious enough for production history and cross-replica locks, and designed so AI workloads and classical services share the same promotion language.

2. The ring model

Rings are shared and ordered (defined once). Each managed application declares, per ring, where it lives and how to reach it — namespace, Deployment, container, image repository, and a health URL. That mapping lives in configuration, not code.

Ring Role Typical AI use
intIntegrationFirst landing for agent/API builds
testTestEval suites, canary traffic, load smoke
accAcceptanceStakeholder / staging gate
prodProductionCustomer-facing AI and APIs

3. Promotion rules (non-negotiable)

  • One ring at a time; never skip. promote {from_ring} always targets the next ring in the pipeline.
  • Source must be healthy before promotion proceeds (live health check).
  • After deploy, the service runs a health check with configurable retries.
  • If the target stays unhealthy after retries, it is automatically rolled back to the previous version.
  • Every seed / promote / rollback is written to history, success or failure.
  • Auto-promote (optional, per ring): a healthy landing can continue onward in the same locked operation — e.g. auto-promote on test so int → test carries on to acc, while acc stays human-gated before prod.

4. Version-verified health (critical for AI)

A plain URL check can be fooled: the deploy “succeeds” but an old inference sidecar or previous agent binary still answers 200 OK. If a ring sets health_version_field (JSON field in the health response, e.g. version or dotted build.version) or health_version_header (e.g. X-App-Version), every post-deploy check also requires the endpoint to report the exact version that was just deployed — otherwise the check fails and auto-rollback kicks in. The same source verifies the source ring really runs the version about to be promoted.

On a ref-pinned ring (ref: release) the expected version is not knowable up front — the pipeline decides what the ref ships — so the field is used the other way: after a healthy deploy the ring records the version the endpoint reports instead of the ref name.

5. Deployers and executors

Clean interfaces keep the promotion engine stable while backends swap:

Concern Production impls
DeployKubectlDeployer, GitHubActionsDeployer, k8sjob
ExecutionGitHub Actions executor; Kubernetes Jobs executor
HealthHTTPChecker (dev: always-healthy fakes)
PersistencePostgres (prod) / memory (dev)
  • KubectlDeployer — shells out to kubectl set image + rollout status, authenticating in-cluster via ServiceAccount.
  • GitHubActionsDeployer — for VM/CI apps that already have a pipeline: triggers workflow-dispatch, waits for conclusion, then applies the same health + rollback logic.
  • k8sjob — runs seed/promote/rollback as a Job in ring-exec; runner env (RP_APP, RP_RING, RP_VERSION, …); stdout streams to step logs; exit code decides success.

The deployer is selected per application, so one control plane can promote Kubernetes apps and VM/CI apps side by side — useful when AI training jobs stay on VMs while serving runs on Kubernetes.

6. Concurrency and reliability

  • Operations on the same application are serialized by a store lock. Postgres uses a session advisory lock, so serialization holds across replicas.
  • Seed/promote/rollback runs under a context detached from the HTTP request and bounded by operation_timeout — a client disconnect cannot abort an in-flight deploy or its automatic rollback.
  • Stored state updates as soon as a deploy lands, so it never lags the cluster even if a later health check and rollback both fail.

7. Why Workstation ships Ring Promoter for AI

Workstation builds AI workstations, private AI, agent platforms, and edge products (including Workstation WSL Proxy). Those stacks need a promotion protocol that treats “AI-powered deployments” as first-class: many apps, mixed deployers, version-honest health, and human gates where they matter. Ring Promoter is that protocol — open at ringpromoter.com and integrated into Workstation solution conversations as the modern CI/CD control plane you cannot miss.

Operators / CI / Agents
    → Ring Promoter (UI + REST API)
        → seed / promote / rollback
        → Deployer (kubectl | GitHub Actions | k8sjob)
        → Health (HTTP + optional version verify)
        → Store (Postgres history + locks)
Rings: int → test → acc → prod

8. Getting started

  1. Visit https://www.ringpromoter.com/ for product context and demos.
  2. Clone github.com/bwalia/ring-promoter and run locally or on k3s.
  3. Read the Workstation product page and the blog summary.
  4. Talk to Workstation about CI/CD for AI platforms via contact.

Published by Workstation. Upstream docs: README and design notes in the Ring Promoter repository.

Product site: https://www.ringpromoter.com/

Diesen Artikel teilen

Mehr zu Technology

Laptops für KI-Entwickler: LLMs lokal unter Linux mit GPUs mit 16 GB und 24 GB betreiben (Leitfaden 2026)

Laptops für KI-Entwickler: LLMs lokal unter Linux mit GPUs mit 16 GB und 24 GB betreiben (Leitfaden 2026)

Was auf eine RTX 5080 oder RTX 5090 Laptop GPU passt, die Speicherrechnung, die lohnenden offenen Modelle, die Einrichtung von Ubuntu 26.04 mit dem NVIDIA-Stack, Ollama, llama.cpp und vLLM, Coding-Assistenten, LoRA-Fine-Tuning, Sicherheit und wann Arbeit auf einen Server gehört

Weiterlesen
ML- und LLM-Systeme testen, deren Korrektheit sich nicht beweisen lässt: ein Python-Evidenz-Kit

ML- und LLM-Systeme testen, deren Korrektheit sich nicht beweisen lässt: ein Python-Evidenz-Kit

Evidenz statt Beweise, in etwa 400 Zeilen Python mit echter Ausgabe: Slice-Metriken mit Bootstrap-Intervallen und einem Pass/Fail-Vertrag, Property- und metamorphe Tests mit Hypothesis, Kalibrierung und Conformal Prediction, Pydantic-Verträge und Regressionsevaluierungen für LLM-Ausgaben, Drift-Prüfungen, ein CI-Gate und die exakte Verifikation eines kleinen neuronalen Netzes

Weiterlesen
TorchCurves in der Praxis: lernbare B-Spline- und Legendre-Kurven in PyTorch

TorchCurves in der Praxis: lernbare B-Spline- und Legendre-Kurven in PyTorch

torchcurves installieren, eine Kurve anpassen und dimensionieren, B-Spline oder Legendre und die Eingabeabbildung wählen, eine monotone Kalibrierungskurve und ein additives Modell auf tabellarischen Daten bauen, mit SciPy, scikit-learn, einem KAN und einem MLP vergleichen, Koeffizienten nach JSON exportieren, testen und in der Produktion betreiben, mit echten Ausgaben

Weiterlesen