--- title: '2022 01 26 ＡＷＳ雲端科技發表會' disqus: hackmd --- ###### tags: `AWS` AWS === [TOC] # 雲端資料庫新功能搶先看 ## Section 1 ### Amazon RDS Multi-AZ DB Cluster ### Amazon RDS Custom ・Focus resources on more strategic tasks to improve your business ・Continue using the same skill sets ・Do not need to rearchitect your application ・Use a simple path to a managed service in the cloud for DB workloads ### Amazon RDS MySQL 8 aurora mysql3 = mysql ㆍA named temporary result set. ㆍCreated using wITH clause ㆍUseful for hierarchical data traversal, avoiding repeated sub-queries, better readability ㆍExists within the scope of a single statement and can be referred within that scope ㆍCan refer to itself or other CTEs. Recursive CTEs when self-referred. #### Role-based access control ㆍ Role is a named container of privileges. ㆍRole support added starting AMS 3 (not available in previous versions) ㆍCreate and drop roles, and grant and revoke privileges to the roles. ㆍStreamlines granting and revoking of privileges to groups of users and simplifies user management. ㆍCan set/change roles during a session using SET ROLE ㆍ SHOW GRANTS to show permissionsRole-based access control ㆍ Role is a named container of privileges. ㆍRole support added starting AMS 3 (not available in previous versions) ㆍCreate and drop roles, and grant and revoke privileges to the roles. ㆍStreamlines granting and revoking of privileges to groups of users and simplifies user management. ㆍCan set/change roles during a session using SET ROLE ㆍ SHOW GRANTS to show permissions #### New index types ㆍNew index types - Invisible, Descending, Functional key parts ㆍInvisible index - ㆍALTER TABLE ... INVISIBLE/VISIBLE ㆍ Optimizer ignores it, but maintained normally ㆍUseful for schema optimization ㆍDescending index - ㆍCreated using the DESC keyword during definition ㆍValues are stored in a descending sort ㆍFunctional index- ㆍIndex expression values rather than column values ㆍEnable indexing of values that cannot be indexed otherwise. ### AWS DMS Fleet Advisor 分析地端資料庫狀況，幫助資料庫遷移 ㆍ Discover and inventory ㆍ Collect metadata ㆍ Gather usage data ㆍ Analyze infrastructure ㆍ Identify migration candidates ㆍ Develop a business case ㆍ Fully integrated into AWS DMS ㆍ Analyze data to develop a roadmap for the cloud journey ㆍ Review infrastructure optimization opportunities ㆍ Discover all possible options to run workloads on AWS ㆍ Build financial models for these possible options ㆍ Make informed choices according to business needs and goals ## Section 2 ### Amazon DynamoDB Standard-IA 計價用存取次數 及 空間 ### Amazon Timestream #### Multi-measure records 可以節省成本 #### Scheduled queries Orders of magnitude faster load times for dashboards and analytical reports, at a fraction of the cost Pre-compute results by storing aggregates, rollups ㆍ Define the query to pre-compute ㆍ Provide a target Timestream table to store the results ㆍ Specify a refresh schedule to update the results ㆍ Pre-computation supports the full SQL surface area of Timestream Fully managed, serverless, scalable, highly available #### magnetic store writes 一樣的API 自動把資料導入到適當的地方 ㆍ Allow insert or updates of data in magnetic store Scenarios: ㆍ Late arrival: data arriving days, months, or years past its associated timestamp - Disconnected devices, historical data ingest ㆍ Data corrections; update older data - Application logic updates, data errors Same write API for memory and magnetic store ㆍ Timestream automatically redirects to appropriate store ㆍ Magnetic store writes applied asynchronously ㆍ Optimize storage costs and enable newer scenarios ### Babelfish for Amazon Aurora PostgreSQl Babelfish ＞ SQL sever 搬移加速器 降低搬遷風險 不須改寫SQL sever原本開發的code ＰＯＣ加速80% #### 支援的 SQL Server-Specific features ㆍTriggers and Stored Procedures ㆍ Nested transaction ㆍTable Valued Functions ㆍData types (money, sql_variant) ㆍSavepoints ㆍStatic cursors ㆍSession Temporary Table ㆍ USE <database> # 現代化無伺服器資料湖倉儲 ## Why modernize 企業希望資料可以在組織內共用 Store all your data in open formats Cost effectively scale storage to exabytes Decouple storage from compute Choice of analytical and ML engines Process data in place ## Modern data architecture ![Imgur](https://i.imgur.com/i3rqgIo.png) ### AWS Glue ![Imgur](https://i.imgur.com/zevx2IQ.png) # 建構從零到全球化的網路架構戰略 ![Imgur](https://i.imgur.com/DFoJJLv.png) # 使用 Amazon DevOps Guru 幫助 AWS RDS 問題監測 ## Amazon DevOps Guru for RDS 偷窺是必然的雖然沒有違反保密協定 ㆍ Finds database performance anomalies ㆍ Analyzes the anomaly ㆍHighlights ㆍ Prevalent wait events ㆍ Prevalent SQL statements ㆍ Other anomalous metrics ㆍ Recommends next steps ## What can it do? - How does it work? - How do you set it up? - How did we build it? ## Example findings