使用AWS Redshift从AWS MSK中读取数据

Amazon Redshift 流式摄取的目的是简化将流式数据直接从流式服务摄取到 Amazon Redshift 或 Amazon Redshift Serverless 的过程。

官方文档[1]中有详细步骤。用unauthenticated, IAM 的方式均可以进行连接,只不过使用的是不同端口:9092或者9098

1\] [https://docs.amazonaws.cn/redshift/latest/dg/materialized-view-streaming-ingestion-getting-started-MSK.html](https://docs.amazonaws.cn/redshift/latest/dg/materialized-view-streaming-ingestion-getting-started-MSK.html "https://docs.amazonaws.cn/redshift/latest/dg/materialized-view-streaming-ingestion-getting-started-MSK.html") 使用IAM方式: dev=# CREATE EXTERNAL SCHEMA external_schema_name_msk FROM MSK IAM_ROLE default AUTHENTICATION iam URI 'b-3.testtest.6ruea7.c2.kafka.cn-northwest-1.amazonaws.com.cn:9098'; CREATE SCHEMA Time: 309.537 ms dev=# SELECT schemaname, tablename FROM SVV_EXTERNAL_TABLES WHERE schemaname = 'external_schema_name_msk'; schemaname | tablename --------------------------+--------------------- external_schema_name_msk | __amazon_msk_canary external_schema_name_msk | __consumer_offsets external_schema_name_msk | kafka-default-topic (3 rows) dev=# CREATE MATERIALIZED VIEW common_msk_cdp_wecom_tag AUTO REFRESH YES AS SELECT * FROM external_schema_name_msk."kafka-default-topic"; CREATE MATERIALIZED VIEW Time: 667.501 ms dev=# select kafka_timestamp from common_msk_cdp_wecom_tag; kafka_timestamp ----------------- (0 rows) Time: 18.564 ms dev=# refresh MATERIALIZED VIEW common_msk_cdp_wecom_tag; INFO: Materialized view common_msk_cdp_wecom_tag was incrementally updated successfully. Stream returned no new data. REFRESH Time: 1493.006 ms dev=# select kafka_timestamp from common_msk_cdp_wecom_tag; kafka_timestamp ------------------------- 2024-10-12 04:49:19.693 2024-10-12 04:49:20.136 2024-10-12 04:49:20.193 2024-10-12 04:49:20.206 2024-10-12 04:49:20.214 2024-10-12 04:49:20.241 2024-10-12 04:49:20.248 2024-10-12 04:49:20.256 (8 rows) Time: 6141.639 ms 使用Unauthenticated 方式: dev=# CREATE EXTERNAL SCHEMA msktest dev-# FROM MSK dev-# IAM_ROLE 'arn:aws-cn:iam::123456789:role/shiranredshift' dev-# AUTHENTICATION none dev-# CLUSTER_ARN 'arn:aws-cn:kafka:cn-north-1:123456789:cluster/shiranmsk/03260cbd-b23c-442f-a858-8af96e6b90de-2'; CREATE SCHEMA dev=# dev=# dev=# CREATE MATERIALIZED VIEW mskview AUTO REFRESH YES AS dev-# SELECT * dev-# FROM msktest."mirror1"; CREATE MATERIALIZED VIEW dev=# dev=# dev=# select * from mskview; kafka_partition | kafka_offset | kafka_timestamp_type | kafka_timestamp | kafka_key | kafka_value | kafka_head ers | refresh_time -----------------+--------------+----------------------+-----------------+-----------+-------------+----------- ----+-------------- (0 rows) dev=#

相关推荐
RestCloud1 天前
SQL Server到Hive:批处理ETL性能提升30%的实战经验
数据库·api
RestCloud1 天前
为什么说零代码 ETL 是未来趋势?
数据库·api
ClouGence1 天前
CloudCanal + Paimon + SelectDB 从 0 到 1 构建实时湖仓
数据库
DemonAvenger2 天前
NoSQL与MySQL混合架构设计:从入门到实战的最佳实践
数据库·mysql·性能优化
AAA修煤气灶刘哥2 天前
后端人速藏!数据库PD建模避坑指南
数据库·后端·mysql
RestCloud2 天前
揭秘 CDC 技术:让数据库同步快人一步
数据库·api
容器魔方2 天前
Bloomberg 正式加入 Karmada 用户组!
云原生·容器·云计算
得物技术2 天前
MySQL单表为何别超2000万行?揭秘B+树与16KB页的生死博弈|得物技术
数据库·后端·mysql
可涵不会debug3 天前
【IoTDB】时序数据库选型指南:工业大数据场景下的技术突围
数据库·时序数据库
ByteBlossom3 天前
MySQL 面试场景题之如何处理 BLOB 和CLOB 数据类型?
数据库·mysql·面试