I would like to know what would be the memory requirements for my data set, I currently have 50GB of data on the whole. And planning to have 6-7 data nodes and 3 master only nodes and will connect using Transport client Java.
What should be my memory requirements, will 50 GB of index will be shared across shards or each data node will be having 50 GB of data? Planning to have default 5 Shards and 1or 2 replica's?
What is the RAM i need and disk-space should i need for each data node to start with?
Since i am planning to have 5 shards and 1 replica, i can go upto 10 nodes , so thought of having 3 master only and then 5 or 6 data nodes. is my assumption wrong, or for the given data size, shall i start with only 3 data nodes?
what memory options would be optimal ?
Thanks, Yes i need dedicated master's and in future i might keep adding more data nodes.
But my original question was regarding memory requirements , can you let me know for 50GB of data what should anyone requirements should be approximately.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.