Hacker News new | ask | show | jobs
by jnaour 3885 days ago
Anyone has experience with this project? We are currently thinking about using it in our cluster with Spark Streaming instead of an external DB (MongoDB)
1 comments

we use it, but (after reading the homepage linked to at the bottom) pretty clearly outside the core use case. In particular, we use it for off-heap caching, which we configure through org.apache.spark.StorageLevel. We monitor via tachyoneFolderName which as you probably know is a spark context (sc) attribute (eg, ./bin/spark-shell, then type "sc."[Tab] to find it).