Creating a noSql database, what is the best source code to look at?

I have always wanted a nosql database that was purpose built for storing large volumes of nested/threaded comments. Implementation would probably be done in java because that is what I am best at. I really like how ElasticSearch is dead simple to set up a cluster and throw data into it, I want my product to share those same qualities. Here are the features I have in mind: 1) auto/manual sharding across clusters 2) auto/manual indexing across clusters 3) full text search (probably via lucene or elasticSearch) 4) REST/JSON API 5) retrieve any comment by ID 6) comments can be retrieved with or without child nodes 7) comment trees can be retrieved with a specified depth 8) comment tree can be retrieved can be filtered by time or rank 9) entire comment trees can be re-parented. What I'm looking for are exceptional pieces of code or specific algorithms that I can study before digging into this project. Can anyone suggest a few places to get started? java  rest  nosql  comments  sharding share improve this question asked  Aug 8 '12 at 3:24 bostonBob 370 4 12

1   Much of this will be a feature of the app you build rather than the database you use. The rest (possibly excluding full-text search), any existing NoSQL database should be able to handle. Why exactly can't you use an already existing DB? –   cHao  Aug 8 '12 at 3:30  Do you want to write your own, or do you want to use one that you like and that is written in Java? –   Edmon Aug 8 '12 at 3:30 About 80% of the reason for wanting to write my own is for fun, the other 20% is because I have never really been fully satisfied with the traditional solutions for storing nested comments. I think it would be cool to be able to fire up a cluster to store/search reddit scale volumes of comments. –   bostonBob  Aug 8 '12 at 3:52 add a comment

1 Answer

active oldest votes

up vote 0 down vote

Since your tag in a question indicates Java, I suggest looking into OrientDB. Here is a source code: http://code.google.com/p/orient/source/browse/ and the architecture: http://code.google.com/p/orient/wiki/Presentations for the big boy stuff (clustering, hyper scaling take a look at HBase and Accumulo): http://hbase.apache.org/source-repository.html http://accumulo.apache.org/source.html Hope this helps. Edmon share improve this answer