A
A
Alexander Kurakin2016-04-18 21:24:40
NoSQL
Alexander Kurakin, 2016-04-18 21:24:40

Which database to choose under these conditions?

Which database to choose under these conditions?

  • open source,
  • non-relational,
  • data is stored and changed on the master, but slaves can cache individual tables,
  • rare writing, frequent reading,
  • tens of columns, hundreds of millions of rows, values ​​- either a number, or text up to a kilobyte, or text up to ten kilobytes.
  • (preferably) column-based,
  • (preferably) MapReduce.

Thank you.

Answer the question

In order to leave comments, you need to log in

1 answer(s)
P
Philipp, 2016-04-19
@kuraga333

Cassandra
Satisfies most of your requirements - column-based, MapReduce, fault-tolerant architecture, ability to work with huge arrays of data. The sparse entry fits well into the column-oriented database model.
You can read this article kkovacs.eu/cassandra-vs-mongodb-vs-couchdb-vs-redis
I have personal experience with MongoDB (3.0, WiredTiger) in the amount of 200+GB and ~20 million rows in a collection with quite a high the number of write / read. So far, even without sharding, we manage.
You can also look towards CouchDB, although these solutions are quite different db-engines.com/en/system/CouchDB%3BCassandra%3BMongoDB

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question