bioinformaticsdatabasesoftware

NoSQL Databases

Originally developed for large-scale web applications, nosql database management system (dbs) like to see themselves as next generation databases and are using “not only sql“. Moving away from the strict model and query approach of traditional , some nosql dbs offer more flexibility and easier scalability. They can be described as being non-relational, distributed, open-source, horizontally scalable, schema-free, easy replication support, simple API, eventually consistent.

There are many different types of nosql dbs being developed, here are some examples:

  1. Document store dbs
    Systems like Apache’s Couch DB store unstructured text and create views on them. The data is stored in JSON and binary formats in a distributed manner, queries often use reduce operations.
  2. Graph dbs
    Neo4j and FlockDB are example systems that store data as nodes with relationships between them.
  3. Key-Value / tuple store dbs
    An even simpler approach is taken by in-memory or on-disk key-value dbs like Redis where simple text or string sets are stored and queried e.g. via a web service.

To try your hands on an example system, you could e.g. use mongodb. This is a type of document-store dbs with the data stored in JSON-like files. The installation on an Ubuntu 16.4 system would follow the instructions here:

sudo apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv 0C49F3730359A14518585931BC711F9BA15703C6
echo "deb [ arch=amd64,arm64 ] http://repo.mongodb.org/apt/ubuntu xenial/mongodb-org/3.4 multiverse" | sudo tee /etc/apt/sources.list.d/mongodb-org-3.4.list
sudo apt-get update
sudo apt-get install -y mongodb-org
sudo service mongod start

By default the service will start on port 27017. To try out some data operations:

mongo
> use no_sql_db_test
> db.person.insert({"name":"Andrea","age":19})
> db.person.insert({"name":"Paul","age":23,"pet":"dog"})
> db.person.find()
  { "_id" : ObjectId("5a12dcdf81636f8f3070ce13"), "name" : "Andrea", "age" : 19 }
  { "_id" : ObjectId("5a12dd0181636f8f3070ce14"), "name" : "Paul", "age" : 23, "pet" : "dog" }
> db.person.find({"age":{$lt:20}})
  { "_id" : ObjectId("5a12dcdf81636f8f3070ce13"), "name" : "Andrea", "age" : 19 }

An overview and intro tutorial can be found here. To stop the db service run:

sudo service mongod stop

Sources & further reading:
c’t (German), Wikipedia, list of systems