4 ms·
Broadcast Hash Joins in Apache Spark SQL
- alexnewman 9y agoThis was the next logical step on a distributed database I worked on back in the day. Good to see someone doing it right
- suj1th 9y agowhich distributed database was it? Just curious.
- alexnewman 9y agoI’ve hacked on Hbase,c5 and small patches to a few others. The one I’m referring to is spite and it was made by drawn to scale. We mapped sql queries to Kahn process networks. We used Lucene for secondary networks but we looked at hash Joins for performance. All the rage in the 70s