Skip to main content
Home
  • About
  • Blog
  • Open Source
  • Resume

Rust

Preview for Pay Once, Not Per Task: Four Repeated Costs Comet 1.1.0 Stopped Paying
Oct 7, 2026

Pay Once, Not Per Task: Four Repeated Costs Comet 1.1.0 Stopped Paying

Apache DataFusion Comet 1.1.0 is out, with 14 of my PRs in it. Four of them fix the same kind of waste: work that gives the same answer every time but was being redone for every task, batch or shuffle block. Here is what each one changed, what the benchmarks really show, and the correctness fix I'd point to first.

apache spark datafusion comet rust scala data engineering performance open source
Read more
Preview for Comet Falls Back to Spark on Every Delta Table. I Wrote the PR That Fixes It
Aug 17, 2026

Comet Falls Back to Spark on Every Delta Table. I Wrote the PR That Fixes It

DataFusion Comet declines every Delta table over a one-line format check, giving up native execution and all parquet pruning. PR #5365 adds a native Delta read path that inherits row-group and page-index pruning for free, and applies deletion vectors inside the scan itself. Measured at 1 TB: the Delta arm is indistinguishable from raw parquet.

apache spark datafusion comet delta lake parquet rust data engineering performance open source
Read more

No posts found matching your search.

Contact

Dustin.William.Smith@gmail.com

Location

Ha Noi, Viet Nam
Creative Commons CC-BY
2026 Dustin Smith
Menu
Home About Blog Open Source Resume