alamb commented on issue #6782: URL: https://github.com/apache/arrow-datafusion/issues/6782#issuecomment-1767075218
Update here: I went over the clickbench results and they look pretty good (details below). I am not sure how much the h2o.ai benchmark is going to add so I am thinking we may just skip that one unless it tell us something useful (it is only operating in 500MB of csv data). cc @JayjeetAtGithub My next steps: 1. Analyze TPCH findings so there is an explanation for the query performance 2. Analyze scalability plots, and explain / reproduce some of the strange scale behaviors for DuckDB (not getting faster on Q37, Q38, etc) 3. Rerun the clickbench scalability on the large core count machine 4. Write up scalability plots on the larger core machine 5. If I have time after this, I will look into h20.ai benchmarks -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
