Configure, Measure, And Optimize Hybrid Tables
Ғылым және технология
This session walks through the key use cases for Hybrid Tables and discusses some real-world challenges, including how to address low latency requirements through its support of indexes, constraints, and row-level locking. It also explores how to ensure performance for hybrid transactional/analytical workloads and covers data modeling considerations, query optimization, and query profiling techniques. It delves into monitoring performance, troubleshooting, and debugging. The session concludes with the sharing of best practices for optimal configurations, resource optimization, understanding boundaries and safe limits, and tuning knobs to optimize and get the best out of Unistore for your workloads.
Subscribe for more! www.snowflake.com/YTsubscribe/
Explore sample code, download tools, and connect with peers: developers.snowflake.com/
Пікірлер: 2
Interesting video - any news on Azure availability?
1. i understand SELECTs scan mode Row or Columnar (makes sense). And how about DML? Can for example a MERGE be executed by either row engine vs native micro partition engine? Same for all sorts of DML (INSERTs, UPDATEs and DELETEs) ? or regarding DML there's only one engine to execute? i have an old feeling on hybrid tables that would first write row based first, and that background process would then copy row based data into micro partitions (yo support typical analytic queries), but while SERVING an analytic query snowflake would merge data from both row based + micro partitions as micro partitions may be lagging till background process row based into micro partitions ? Thanks 🙏