I have just started exploring Elastic Search. I have a requirement similar to below thread
I want to store the data specific to brands for efficient search. I have around 4k brands and each brand has around 10k files. Initially I was thinking of creating one index per brand and store data. But looking at above link it is not efficient.
Second option was to create different index type for each brand. But I read somewhere that number of type should not be more than 5-6.
Files would be pdf,msg,xls,eml etc. I am planning to use ingest attachment plugin to import these files to ES.
File size is 1-20 mb.
In my case the requirement is to search the data related to one brand at a time. Do you think creating only one index and storing all brand related data in the same index would decrease the performance?
Due to your file sizes, it'll create a very big indice if files are all at 20mb. In an other hand, having 4k indices (one per brand) will not be easy to administrate... Anyway, you'll have to store all this datas in one or more indices. Try to create one indice, add documents in, and monitore sizes. Add several nodes and monitore your platform
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.