Before we dive into the installation of Cassandra, let’s first familiarize ourselves with what Cassandra is and its architecture.
What is Cassandra?
Cassandra is a NoSQL distributed database(decentralised) based on peer-peer architecture as it is masterless with high availability, scalability and data distribution across multiple nodes. Cassandra provides a high availability service with no single point of failure.
Architecture of Cassandra
The design of cassandra is done in such a way that it can handle enormous amounts of data across multiple nodes without any single point of failure. Cassandra has a peer-peer distributed system across it’s nodes, and the data is distributed among all the nodes in a cluster.
1) All the nodes in a cluster play the same role. Each node is independent and at the same time interconnected to other nodes.
2) Each node in a cluster can accept read and write requests, regardless of where the data is actually located in the cluster.
3) When a node goes down, read/write requests can be served from other nodes in the network.
In Cassandra, one or more of the nodes in a cluster act as replicas for a given piece of data. If it is detected that some of the nodes responded with an out-of-date value, Cassandra will return the most recent value to the client. After returning the most recent value, Cassandra performs a read repair in the background to update the stale values.
Let’s start, there are three common methods of installing the cassandra i.e.
1) Docker Image
2) Package installation (RPM)
3) Tarball binary file
For Docker image installation, you need Docker to be installed in your system. If you already have it, then you need to pull the cassandra image from dockerhub.
For Package installation in Debian-based distributions(ubuntu), you can install cassandra using APT. Note that to install cassandra using APT requires root permission and a new Cassandra OS user is created along with its binaries and configuration files.
For Tarball binary file installation, download the tar file and extract it. The tarball file extracts all of its contents and its binaries, configuration in the directory respectively. Here, we are going to install the cassandra using tarball binary file.
Installing the cassandra using binary tarball
1) Verify if java is installed in your system. Otherwise install java in your system.
2) Download the tarball file of cassandra from this site and extract it.
The directories after extracting the tarball file is shown:
bin/ conf/ data/ doc/ interface/ javadoc/ lib/ logs/ pylib/ tools/
1) The bin directory consists of commands to run cassandra, nodetool, cqlsh and SSTable tools.
2) The conf directory consists of a cassandra.yaml file in which the configuration is available to start a single node cassandra cluster. For multi-node clusters, the configuration must be modified to required criteria.
3) The data directory consists of information on commit logs, hints and SSTables.
4) The logs directory consists of system and debug logs.
Setting Up Cassandra in Cluster
In our cluster, we are using 4 nodes and setting up cassandra in it. But you can set up any number of clusters in your machine. The process of setting up cassandra is given below:
1) Firstly, Install the cassandra using tarball file in your machine(node) as mentioned above.
2) Secondly, in the conf directory there is cassandra.yaml file which holds the configuration file and loads during startup of cassandra. Add ip address of the nodes which you wanna connect in the seeds field found in cassandra.yaml file.
3) Repeat above steps in all the nodes.
4) At last start the cassandra in every node using the command: bin/cassandra -f which is found in the bin directory of the extracted tarball file of cassandra.
You have successfully installed Cassandra in your system.
Large volumes of unstructured data can be an issue for traditional relational databases. This popular NoSQL database solution is going to allow you to capture and store a lot more of increasingly valuable data.