This week we shipped the infra for automated benchmarks, including:
- Query TSV files without a table function. Raw benchmark data is uploaded as TSV files, so we made it easy to query them directly without having to specify a table function.
- Support for larger file reads in
GlareDB is a light and fast SQL database for analytics
Follow along at glaredb.com/changelog
- This week we introduced support for multi-file reads, metadata columns, faster remote CSV reads, and async filesystem state loading. See what else we shipped this week at glaredb.com/changelog
- Last week we shipped a bunch of performance optimizations and our first set of ClickBench results: github.com/ClickHouse/Cli… This week we expanded object store integration and introduced a unified globbing API that works across S3, GCS, and the local filesystem. Full release
- 🚀 New on the GlareDB blog: SQL Functions in GlareDB: Automatic Vectorization & Type Casting Learn how easy it is to implement SQL functions in GlareDB--our executors handle nulls, casting, and vectorized execution for you. Read more 👉
- New on the GlareDB blog -> Compiling GlareDB to WebAssembly: A Peek Under the Hood. Learn how we built our engine for WASM--handling threading, I/O, and a unified shell--all in Rust. Dive in: glaredb.com/blog/peek-at-w… #Rust #WebAssembly #GlareDB

