I was going through one of the webinars by elastic.In the webinar,I came across a formula to calculate data nodes as below:
Data Nodes: ROUNDUP(Total Storage(GB) / Memory per data node / Memory:data ratio) + 1 Data node for failover capacity
My requirement is as below:
Total Storage: 10,080 GB
Memory per node: 64 GB
Memory:data ratio: 1:30
But when I use above formula(assuming it mentions dividing total storage by memory per data node and then by ratio), my final value for number of data nodes comes out to be 5251 nodes.
Data Nodes: ROUNDUP(10080/64/(1:30))+ 1
I want to know whether I am calculating this correctly or is there any problem with the formula?
@Nikhil04 curious how you came up with the total storage does that include a primary plus one replica of your data or just the primary if so you need to double the total storage so that you can have one primary copy plus one replica copy in order to have HA... l.e. so a loss of a data node does not equal loss of data or make the cluster unusable
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.