Search code, repositories, users, issues, pull requests...
KubeBlocks is an open-source control plane software that runs and manages databases, message queues and other data infrastructure on K8s. The name KubeBlocks is inspired by Kubernetes and LEGO blocks, signifying that running and managing data infrastructure on K8s can be standard and productive, like playing with LEGO blocks.
KubeBlocks could manage various type of engines, including RDBMSs (MySQL, PostgreSQL), Caches(Redis), NoSQLs (MongoDB), MQs(Kafka, Pulsar), and vector databases(Milvus, Qdrant, Weaviate), and the community is actively integrating more types of engines into KubeBlocks. Currently it has supported 32 types of engines!
The core of KubeBlocks is a K8s operator, which defines a set of CRDs to abstract the common attributes of various engines. KubeBlocks helps developers, SREs, and platform engineers deploy and maintain dedicated DBPaaS, and supports both public cloud vendors and on-premise environments.
Kubernetes has become the de facto standard for container orchestration. It manages an ever-increasing number of stateless workloads with the scalability and availability provided by ReplicaSet and the rollout and rollback capabilities provided by Deployment. However, managing stateful workloads poses great challenges for Kubernetes. Although StatefulSet provides stable persistent storage and unique network identifiers, these abilities are far from enough for complex stateful workloads.
