For uninterrupted computing services
This guide walks through preparing the Kimi 2.5 model, uploading it to OSS, configuring persistent storage, and using SGLang, RoleBasedGroup, and Openclaw to deploy a production‑grade inference service on Alibaba Cloud Kubernetes with step‑by‑step commands and YAML examples.
Kimi built a high‑performance, low‑cost AI Agent infrastructure by combining Alibaba Cloud ACK node pools and the ACS Agent Sandbox, addressing challenges of instant sandbox response, state continuity, massive concurrency, cost efficiency, security isolation, and search‑memory integration for production‑grade agents.
This article details how to deploy the 235‑billion‑parameter Qwen3‑235B model using PD‑separation and MoE techniques, explains the associated challenges, and demonstrates a production‑grade solution built on the high‑performance SGLang inference engine and the RoleBasedGroup (RBG) orchestration framework, complete with benchmark results and best‑practice YAML examples.
This guide explains how to set up Alibaba Cloud Service Mesh (ASM) on an ACK Kubernetes cluster, covering prerequisites, two methods of cluster registration, creation of north‑south and east‑west gateways, traffic routing with HTTPRoute, security policies using PeerAuthentication and AuthorizationPolicy, and observability configuration via Telemetry.
This guide explains how to combine AgentScope with Knative to efficiently develop, build, and deploy AI agents using serverless containers, covering key features, pain‑point solutions, installation steps, code examples, deployment commands, and post‑deployment observations.
The article explains how OpenKruise Agents, an open‑source project from Alibaba Cloud, provides a cloud‑native sandbox infrastructure for AI agents on Kubernetes, detailing its architecture, lifecycle management, security challenges, resource pooling, and future roadmap for AI‑driven workloads.
This article explains why AI inference services require multi‑cluster gray‑release, outlines the risks of traditional updates, and details how ACK One Fleet combined with Kruise Rollout provides a controlled, observable, and rollback‑capable solution for deploying large AI models across hybrid cloud clusters.
This article explains how the cloud‑native Knative serverless framework reduces GPU waste, enables request‑driven autoscaling to zero, accelerates AI model versioning and startup with Fluid, and integrates protocols like MCP and A2A to deliver cost‑effective, high‑performance AI inference services.
This article explains why KV‑Cache hit rate is critical for large‑model inference, describes vLLM's automatic prefix caching, outlines the distributed cache challenges, and provides a step‑by‑step guide to deploying Alibaba Cloud ACK Gateway with Inference Extension's precise‑mode prefix‑cache‑aware routing, backed by benchmark results.
At KubeCon NA, experts showed that combining deterministic Argo Workflows with large‑model AI agents lets teams orchestrate smart, flexible agents in a predictable, observable, and auditable way, enabling large‑scale CVE remediation and self‑healing operations on Kubernetes.
