Skip to content

Latest commit

 

History

History
171 lines (119 loc) · 8.23 KB

README.md

File metadata and controls

171 lines (119 loc) · 8.23 KB
Raw

Anyscale Stack

Anyscale is an enterprise Core Data & Core AI stream analysis platfrom, the full name of Anyscale is Anyscale Stack，Analyze flow data . Anyscale is not a framework, but a complete solution, mainly includes：

  • Basic concept: Anyscale based on Streaming Analytics Visualization architecture

  • Development tools: Ambari, based on the Ambari source extension, Anyscale is better integrated with Ambari.

  • The basic structure: Anyscale core is based on high-performance distributed relational database construction, supporting triceratops of data-second responsiveness.

  • Core features: Anyscale Core Data & Core AI:

    • Core Data provides highly visual stream data analysis.
    • Core AI provides AI capabilities on distributed systems.

  • Basic operation and maintenance: out of the box, visual operation and maintenance management system Ambari Manager:

    • Provide core service management and configuration center functions to ensure the stability and reliability of the cluster.
    • Manage large-scale distributed systems simply and efficiently, just need to pay attention on business applications.

Anyscale

Design Principles

  • Internet of Things(IoT), connects everything’s era.Data is processed in the flow
  • Anyscale flow analysis platform, simple and high performance data analysis, Core Data follows the principle of distributed system design
  • Follow the existing service system of company. seamlessly blend into existing enterprise data center systems that could directly interact with data.

features

  • One-stop solution: Out of the box & simple & easy to use.
  • High performance: teraflops of data-second response, providing PB-level data storage.
  • Scalability : Large-scale deployment, up to several hundred clusters.
  • Visualization: Visual data collection, data storage, data analysis, data reporting.
  • Reliability: The cluster provides a copy-tolerant mechanism, so hardware failures do not cause data loss.
  • Simple operation and maintenance: log, monitoring, alarm, configuration, and service stack management.
  • Easy to use: Provides standard SQL, drag and drop data analysis.

Anyscale Development plan for 3.0.0.0

  • Project List
Name explain address status remark
Anyscale-package Anyscale Ecosystem Package Management Tools Github ok 🌱
Anyscale Anyscale for roadmap Github ok 🌱
Anyscale-select Anyscale Ecosystem Package Core Tools Github ok 🌱
Anyscale-ambari-mpack Anyscale Ecosystem Management Platform Github ok 🌱
Ambari Anyscale Stack in Ambari Github ok 🌱

Anyscale Development plan for 3.1.0.0

  • Project List
Name explain address status remark
anyscale-package Anyscale Ecosystem Package Management Tools Github ok 🌱
anyscale-select Anyscale Ecosystem Package Core Tools Github ok 🌱
ambari-plus Anyscale Stack in Ambari Github ok 🌱

Note: The Anyscale-ambari-mpack and Anyscale-ambari projects will be merged in 3.1.0.0, and only one project will exist in the future.

Anyscale Development plan for 3.1.1.0

R & D Plan

  • Add Hadoop 3.1.1
  • Add Spark 2.4.0
  • Add Hbase 2.0.2
  • Add Livy 0.5.0
  • Add Flink 1.7.0

Test Plan

  • Add Hadoop 3.1.1 (done)
  • Add Spark 2.4.0 (done)
  • Add Hbase 2.0.2 (done)
  • Add Livy 0.5.0 (done)
  • Add Flink 1.7.0 (Not tested)

Project List

Name explain address status remark
Anyscale-package Anyscale Ecosystem Package Management Tools Github release 🌱
Anyscale-select Anyscale Ecosystem Package Core Tools Github release 🌱
Ambari Anyscale Stack in Ambari Github release 🌱
FusionDB Anyscale Core FusionDB Github incubator 🌱

Note: Anyscale 3.1.1.0 version, incubator core components FusionDB and CoreAI，details access fusionlab

Anyscale Development plan for 3.2.0.0

R & D Plan

  • Add FusionDB 0.1.0
  • Add CoreAI 0.1.0

Test Plan

  • Add FusionDB 0.1.0
  • Add CoreAI 0.1.0

Project List

Name explain address status remark
Anyscale-package Anyscale Ecosystem Package Management Tools Github release 🌱
Anyscale-select Anyscale Ecosystem Package Core Tools Github release 🌱
Ambari Anyscale Stack in Ambari Github release 🌱
FusionDB Anyscale Core FusionDB Github incubator 🌱

Note: Anyscale 3.2.0.0 version, release core components FusionDB and CoreAI，details access fusionlab

Anysacle Development plan for 1.0.0.0 (restart: changed brand to Anyscale stack)

Plan

  • Ambari-plus 2.7.4.0
  • FusionDB 1.0.0.0
  • Hadoop 3.1.1
  • Zookeeper 3.4.6
  • Metastore 3.1.0
  • Spark 3.1.0
  • Hbase 2.1.6
  • CoreAI 0.1.0

Test Plan

  • Ambari-plus 2.7.4.0 (done)
  • FusionDB 1.0.0.0
  • Hadoop 3.1.1 (done)
  • Zookeeper 3.4.6 (done)
  • Metastore 3.1.0
  • Spark 2.4.0 (done)
  • Hbase 2.1.6 (done)
  • CoreAI 0.1.0

Project List

Name explain address status remark
jdp-package Anysacle Ecosystem Package Management Tools Github release 🌱
jdp-select Anysacle Ecosystem Package Core Tools Github release 🌱
Ambari-plus Anysacle Stack in Ambari Github release 🌱
FusionDB Anysacle Core FusionDB Github incubator 🌱

Note: Anysacle 1.0.0.0 version, release core components FusionDB and CoreAI，details access fusionlab

Reporting issues

We use GitHub Issues to track community reported issues. You can also contact the community for getting answers.

Development

项目列表，涉及的相关项目版本，全都统一为branch-3.1.0.0，Anyscale-package除外，拥有相关prefix，suffix需保持一致。 比如：Anyscale-3.1.0.0需具备版本一致性。

需要开启统称为branch-3.1.0.0分支以此为基础进行开发。最终release版本之前，可通branch-3.1.0.0分支发布带rc后缀的版本，比如rc1,rc2等。

新特性开发，可开启带develop关键字的分支，每个roadmap特性都会以pr的形式提到branch-3.1.0.0中，最终版本合并到master，并且打进行tag发布。

  • {project_name}-number(unique)-feature-{describe}

bugfix开发，相关问题修复，需开启带有bugfix-{describe}关键字的分支，测试通过以pr的形式提给branch-3.1.0.0。

  • {project_name}-number(unique)-bugfix-{describe}

commit信息，开发新特性和问题修复，严格以全英文方式提交信息，尽量让每一个pr都能清楚描述解决的问题或新特性。

文档编写

  1. 新特性的开发，需在文档项目中，编写详细的设计文档，设计文档评审通过，开始实现功能。
  2. 每个新特性开发合并branch-3.1.0.0，需要编写相关使用文档，每次提交CI持续集成会自动部署为一个静态网站，只需专注编写文档。整个文档严格遵守markdown语法规范，文档内容描述需通俗易懂，易于理解。

。。。。