co0lster

6 ms·

1. 1.▲
   by co0lster 3mo ago

   duckdb relies on filter pushdowns that makes the reason harder to spot, but you don’t use sqlite to connect to database via JDBC protocol, are you?
2. 2.▲
   by co0lster 3mo ago

   You spawn in memory instance of duckdb and connect to it.
3. 3.▲
   by co0lster 10mo ago

   650GB relates to size of parquet files which are compressed in reality it’s way more. 32 GB of parquet cannot fit in 32GB of RAM
4. 4.▲
   by co0lster 2y ago

   I’m stealing Spotify random instrumental music idea! Why I cannot come up with similar ideas by my own? Thats the hardest part I think…
5. 5.▲
   What's New in Spark 3.3.0(datapotion.io)

   3 points by co0lster 4y ago|0 comments
6. 6.▲
   Visualize StackOverflow Users as a Graph(blog.datahunters.ai)

   1 points by co0lster 5y ago|0 comments
7. 7.▲
   Time Traveling in Data Lakes(billennium.com)

   8 points by co0lster 5y ago|0 comments