query performance

On this episode, we debunk the myth that Druid can't do joins. Druid doesn't function as a traditional relational database because it was purpose-built for lightning-fast queries on large datasets. However, this doesn't mean Druid is entirely devoid of join capabilities – it simply approaches them differently. Our myth-busting team features returning guests Sergio Ferragut and Hellmar Becker from Imply ready to clarify how Druid handles joins in its own unique way and tackle what Druid is for in the first place. 

How do you keep your SQL queries fast? Richard chats with Monica Rathbun about her approaches to SQL Server query tuning. Monica starts with defining the problem - how do we know that the database is the performance bottleneck? The conversation dives into measuring query performance and the power of Query Store, but only on SQL Server 2016 and above, so get upgrading! Entity Framework is a standard tool for developers to automate access to SQL. Still, it can generate some pretty ugly queries, and Monica talks about different ways to improve them, including the old standby of writing a stored procedure. Lots of ideas for folks struggling to make their databases go fast!

Links:

• Monitoring Performance with Query Store
• Common Table Expressions
• Cursors in SQL Server
• SQL Server Profiler
• Glenn's SQL Server Performance Blog

Recorded February 22, 2022