What is this?
Apache Cassandra is an open source, transactional distributed database — a highly-scalable partitioned row store built for linear scalability and proven fault-tolerance on commodity hardware or cloud infrastructure.
Apache Cassandra is an open source, transactional distributed database — a highly-scalable partitioned row store built for linear scalability and proven fault-tolerance on commodity hardware or cloud infrastructure.Is this a server I run, or a library I embed in my app?It is a database server, not a library: you run it as a cluster of nodes and query it over the network with CQL (Cassandra Query Language), for example through the bundled interactive cqlsh client.
It is a database server, not a library: you run it as a cluster of nodes and query it over the network with CQL (Cassandra Query Language), for example through the bundled interactive cqlsh client.How do I get a basic one-node cluster up and running?Unpack the release archive (tar -zxvf apache-cassandra-$VERSION.tar.gz), start the server with 'bin/cassandra -f' (it stays in the foreground and logs to stdout; Ctrl-C stops it), then run 'bin/cqlsh' to create tables and read/write data.
Unpack the release archive (tar -zxvf apache-cassandra-$VERSION.tar.gz), start the server with 'bin/cassandra -f' (it stays in the foreground and logs to stdout; Ctrl-C stops it), then run 'bin/cqlsh' to create tables and read/write data.What do I need installed before starting it?A supported Java version — the exact list is the java.supported property in build.xml — plus Python, which the cqlsh shell requires (its supported versions are checked in bin/cqlsh).
A supported Java version — the exact list is the java.supported property in build.xml — plus Python, which the cqlsh shell requires (its supported versions are checked in bin/cqlsh).How close is CQL to SQL — can I just write my usual SQL?CQL is a close relative of SQL, but think of it as 'SQL minus joins and subqueries, plus collections.' If your workload leans on joins or subqueries, you will have to model your tables around that.
CQL is a close relative of SQL, but think of it as 'SQL minus joins and subqueries, plus collections.' If your workload leans on joins or subqueries, you will have to model your tables around that.What happens to my data when I add or remove machines from the cluster?Cassandra rebalances on its own: it automatically repartitions data as machines are added and removed, and distributes it across nodes transparently to the application.
Cassandra rebalances on its own: it automatically repartitions data as machines are added and removed, and distributes it across nodes transparently to the application.What does a first CQL session actually look like?From cqlsh you create a keyspace with a replication setting (e.g. SimpleStrategy with replication factor 1), then CREATE TABLE users with a userid PRIMARY KEY, INSERT a row and SELECT it back — at that point your one-node cluster is operational.
From cqlsh you create a keyspace with a replication setting (e.g. SimpleStrategy with replication factor 1), then CREATE TABLE users with a userid PRIMARY KEY, INSERT a row and SELECT it back — at that point your one-node cluster is operational.How is the data model organized inside?Like a relational database: rows are organized into tables with a required primary key. The difference is that the store is partitioned, so rows are spread across the machines of the cluster instead of living on one server.
Like a relational database: rows are organized into tables with a required primary key. The difference is that the store is partitioned, so rows are spread across the machines of the cluster instead of living on one server.What are similar projects?scylladb, nebula, tidb, zookeeper, druid, foundationdb, yugabyte-db, trino
scylladb, nebula, tidb, zookeeper, druid, foundationdb, yugabyte-db, trinoWhich topics is it in?cassandra, apache, distributed-database, scalability, big-data, nosql, storage, sql, distributed-systems, cloud, java, backend
cassandra, apache, distributed-database, scalability, big-data, nosql, storage, sql, distributed-systems, cloud, java, backend