Ideas, field notes and deep dives into cloud architecture, integration and AI.
TL;DR : What powers Azure Container Apps Sandboxes? An architectural exploration of microVMs, Cloud Hypervisor, nested virtualization, golden snapshots, and prewarmed pools. The key insight: microVMs provide isolation, but the real innovation is the control plane that turns them into durable execution environments for AI agents.
TL;DR : Keep your integration layer solid and deterministic, and you can let your AI agents run wild with creative problem-solving without breaking the enterprise.
TL;DR : Can you bring a .NET agent into kagent? Absolutely! This article walks through building a custom agent with Microsoft Agent Framework, exposing it via A2A, deploying it to Kubernetes, and integrating it into a multi-agent workflow. A practical demonstration of framework-agnostic, cloud-native agent collaboration.
TL;DR AI hype cooled in 2025, but scaling still depends on data readiness, context engineering and organizational change. A year of community sharing included five conferences, four community events, APIM Hotrod Season 2 and a technical book review, with human judgment central to AI collaboration.
TL;DR AI coding agents shift pull requests from personal ownership toward shared collaboration: humans define intent, steer implementations and judge quality while agents do the coding. The speed gains require balance, avoiding excessive parallel work that creates context switching and cognitive overload.
TL;DR : Stop alerting on metrics. Start alerting on health. Azure Monitor Health Models combine metrics, logs, and dependencies into a unified view of workload health. Model critical business flows, propagate failures across components, and trigger meaningful alerts that reduce noise and reflect real customer impact.
TL;DR Azure Container Apps serverless GPUs offer managed, consumption-based compute with autoscaling to zero. A Hugging Face TGI deployment serves a Mistral model and exports traces to Application Insights, while exploring GPU selection, cold starts, regional availability and scaling limits.
TL;DR Logic Apps Hybrid scales individual pods rather than an entire Workflow Plan. Tests show KEDA and HPA adjusting replicas using SQL-backed job metrics and HTTP concurrency. Manual scaling uses overlapping revisions that need spare capacity, and scaling to zero is not supported.
TL;DR : Discover key insights to successfully run Logic Apps in a hybrid environment.
TL;DR As the year winds down and the busiest period is behind us, I’m taking a moment to reflect on 2024.
TL;DR Logic Apps Hybrid returns on Azure Arc and the Container Apps extension, hosting workflows on your Kubernetes infrastructure with SQL state storage and SMB workflow files. A deployment walkthrough explores local execution, Azure management and preview limits, including required cloud connectivity.
TL;DR AKS Automatic simplifies cluster provisioning, scaling, security and observability, meeting 26 of 33 evaluated high-priority AKS Checklist controls. It provides strong defaults, but production readiness still depends on closing gaps in areas such as registry integration, network policies and alerts.
TL;DR Celebrating ten years of Azure API Management and its evolution into a key platform for securing, monitoring and scaling enterprise APIs. A special APIM Hotrod episode with the Microsoft product group revisits its origins, growth and impact on business innovation.
TL;DR AI gateways centralize governance, security and observability for LLM access instead of duplicating them in each app. Azure API Management demos show priority-based load balancing, RAG request tracing and basic prompt filtering, while highlighting the effort of maintaining custom AI policies.
TL;DR Join APIs in Action 2024 on February 28 at 17:00 CET for three lightning talks.
TL;DR Starting 2024 with the Microsoft MVP Award feels amazing.
TL;DR Migrate VNET-integrated Azure API Management from stv1 to stv2 by comparing in-place and parallel approaches. A practical in-place migration avoids user downtime through network impact analysis, a new subnet and public IP, NSG updates, DNS cutover and infrastructure-as-code changes.
TL;DR A return to blogging in 2023 brought 21 posts, talks at three conferences and one community event, and four Microsoft newsletter features. Azure API Center, Container Apps workload profiles and Cloud for Sovereignty led readership, closing a rewarding year of sharing with the Azure community.
TL;DR Azure Container Apps service discovery resiliency adds retries, timeouts, circuit breakers and connection pools without application code changes or explicit sidecars. A two-service demo tests retry and circuit breaker policies, configured on the called app and used through ACA service discovery.
TL;DR The Karpenter provider for AKS provisions nodes sized to pending pod requirements instead of relying on fixed VM scale set configurations. A preview walkthrough tests dynamic scale-out and consolidation, showing how workload-aware node selection can improve efficiency and cost control.
TL;DR Radius models applications and their dependencies in Bicep, using Kubernetes environments and operator-defined recipes to hide infrastructure details. A first deployment explores its developer experience, portability promise and open questions around maturity, governance and GitOps.
TL;DR Testing Microsoft Defender for APIs with 300 million mirrored requests reveals alerts for traffic spikes, suspicious IPs, unusual payloads and parameter anomalies. The post explains APIM onboarding, learned traffic baselines and data retention considerations for security teams.
TL;DR Microsoft Cloud for Sovereignty helps governments adopt public cloud with a Sovereign Landing Zone, policy guardrails and confidential workload templates. A preview walkthrough covers platform deployment, compliance, remediation and exemptions, emphasizing careful governance.
TL;DR Azure Container Apps workload profiles combine Consumption and dedicated compute in one environment, adding sizing, cost and networking flexibility. Practical tests explore subnet capacity, replica placement and why revision changes need spare resources to avoid degraded deployments.
TL;DR Run scheduled tasks on Azure Container Apps using native Jobs or a continuously running app with a Dapr cron binding. Native Jobs are the simpler, more resource-efficient choice, with built-in execution history, while Dapr requires an active replica and custom log queries.
TL;DR OpenTelemetry standardizes traces, metrics and logs so you can instrument once and send telemetry to multiple backends. A .NET and Azure Service Bus demo shows context propagation and Collector pipelines, decoupling application instrumentation from observability tools.
TL;DR Azure Deployment Stacks group resources across scopes, enforce protection and control deletion or detachment through lifecycle policies. A hands-on walkthrough covers creating, updating and deleting stacks, plus naming and output changes when migrating from deployments.
TL;DR Azure API Center brings API inventory and custom metadata to enterprise governance. An early preview explores dependency tracking with Cosmos DB graphs, API design workflow integration and stakeholder needs, highlighting opportunities for built-in visualization and automation.
TL;DR A hands-on comparison finds no universal winner: Logic Apps offers rapid visual development, rich connectors and mature operations, while Dapr workflows offer code-based flexibility, portability and replay. Choose based on integration scope, skills and operational needs; Dapr was still alpha.
TL;DR Clarifying where Logic Apps Consumption, Standard, Azure Functions, Durable Functions and Dapr fit in integration scenarios. Logic Apps Standard adds networking and extensibility options, but its custom .NET Framework code feature requires a wrapper and Windows hosting.
TL;DR Kubernetes 1.25 introduces cgroup v2, which can cause memory pressure and OOM kills in older .NET and Java applications. Updating frameworks is the lasting fix; reverting AKS nodes to cgroup v1 buys time, but node reimaging during upgrades reinstates v2.
TL;DR Design landing zone monitoring around collection, visualization and alerting requirements. Azure Monitor baseline alerts use policy initiatives to deploy consistent alert rules at scale; existing resources need remediation, and ongoing policy compliance monitoring is essential.
TL;DR Join my Azure Day 2023 Rome livestream session on June 23 at 16:30 CET to learn how Data API Builder, Azure Container Apps and API Management work together to build and secure serverless APIs.
TL;DR The managed Istio add-on lowers the barrier to service mesh adoption on AKS. This walkthrough enables the add-on, configures ingress routing, injects sidecars through namespace labels and adds Zipkin tracing, while noting configuration and upgrade tradeoffs.
TL;DR Build Azure landing zones with Bicep orchestrators, reusable solution modules and Microsoft-maintained CARML building blocks. Bicep registries simplify sharing, but choosing public, customer-owned or partner-managed registries requires balancing security, versioning and maintenance.
TL;DR Run Traefik on Azure Container Apps as an alternative to the APIM self-hosted gateway. Mount static and dynamic configuration from Azure file shares, configure backend routing and chain middleware such as rate limiting to control traffic across containerized APIs.
TL;DR Choose AKS for Kubernetes flexibility, portability and advanced workloads; choose Azure Container Apps for simpler deployment and less infrastructure management. Weigh complexity, scale, operational skills, feature constraints and vendor lock-in against your application needs.
TL;DR Deploy Data API Builder on Azure Container Apps to turn database objects into serverless APIs. Store dab-config.json in an Azure file share, mount it into the DAB container and use Bicep to configure the image, volume mounts and startup arguments.
TL;DR Join my Techorama Belgium session on May 16, 2023 at 13:15 in room 12 to learn how to build and secure serverless APIs with Data API Builder, Azure Container Apps and API Management.
TL;DR Data API Builder turns database objects into REST and GraphQL APIs without code. This walkthrough exposes a SQL table, tests CRUD and filtering, and outlines security controls. Directly exposing a database model creates coupling and security risks, so choose what to expose carefully.
TL;DR Azure Kubernetes Service provides managed Kubernetes on Azure to help teams deploy and manage containerized applications. This visual overview introduces reasons to consider AKS as a container orchestration platform.
TL;DR Containers package an application and its dependencies into a single portable unit, simplifying deployment and management while enabling consistent execution across developer machines, on-premises infrastructure and cloud environments.
No articles match this topic.