Introduction and release notes for Knative components - Container Service for Kubernetes

Knative is an open source Kubernetes-based serverless framework. The purpose of Knative is to create a cloud-native and cross-platform orchestration standard for serverless applications. Knative helps you build an enterprise-class serverless container platform to deploy and manage modern serverless workloads.

Introduction

Container Service for Kubernetes (ACK) Knative is completely compatible with open source Knative and is interfaced with ACK, message services, storage services, and networking services to provide production-level capabilities.

Knative consists of two core components. These components comprise a general-purpose serverless framework:

Knative Serving: manages serverless workloads based on events and can automatically scale pods based on resource requests. If no workload is processed, the number of pods is scaled to zero.
Knative Serving is used to manage the workloads of serverless applications that provide services to users. The most important feature of Knative Serving is pod auto scaling. Pods can be scaled without limits. Knative Serving also supports canary release.
Knative Eventing: provides event management capabilities, such as producing and consuming events.
Knative Eventing provides an all-in-one system for event-driven serverless applications. The system provides features that allow you to install external event sources, register and subscribe to events, and filter events. The event system decouples event producers and event consumers. An event producer can generate events before active event consumers listen to events. An event consumer can listen to events before active producers produce events.

Usage notes

For more information about how to use Knative, see Deploy Knative. For more information about Knative, see Knative overview.

Release notes

December 2023

Version	Release date	Description	Impact
v1.10.6-aliyun.1	2023-12-25	Existing Application Load Balancer (ALB) instances can be reused. The downward API is supported by Knative Services. The CVE-2023-48713 vulnerability is fixed. Important The Kubernetes version of the cluster must be 1.24 or later.	No impact on workloads. We recommend that you update Knative during off-peak hours.

October 2023

Version	Release date	Description	Impact
v1.8.3-aliyun.1	2023-10-17	Knative domain mappings can be used to configure Microservices Engine (MSE) gateways to access Services over HTTPS. The issue that instances are reserved during a scale-in activity is fixed.	No impact on workloads. We recommend that you update Knative during off-peak hours.

July 2023

Version	Release date	Description	Impact
v1.8.2-aliyun.1	2023-07-20	Important The Knative version must be 1.24 or later. Pods are created based on the `autoscaling.knative.dev/activation-scale` setting to handle traffic spikes when Knative initializes. Knative 1.4.x can be updated to 1.8.2.	No impact on workloads. We recommend that you update Knative during off-peak hours.

June 2023

Version

Release date

Description

Impact

v1.4.5-aliyun.1

2023-06-19

Pods can be scaled to zero by ACK Serverless Knative, which works based on MSE gateways.

KServe is supported.

No impact on workloads. We recommend that you perform the update during off-peak hours.

May 2023

Version	Release date	Description	Impact
v1.4.3-aliyun.1	2023-05-18	MSE gateways are supported. Precise scaling control based on request concurrency is supported.	No impact on workloads. We recommend that you update Knative during off-peak hours.

April 2023

Version	Release date	Description	Impact
v1.4.2-aliyun.1	2023-04-26	Kubernetes 1.26 is supported. GPU metrics-based predictive scaling by using Advanced Horizontal Pod Autoscaler (AHPA) is supported. The Kourier gateway version issue is fixed.	No impact on workloads. We recommend that you update Knative during off-peak hours.

March 2023

Version	Release date	Description	Impact
v1.4.1-aliyun.1	2023-03-23	Reserved instances are supported. The podspec-fieldref parameter is enabled by default. Startup issues in Kafka sources are fixed.	No impact on workloads. We recommend that you perform the update during off-peak hours.