February 2012 Summaries

The author has announced a partnership with Pentaho, providing an easy-to-use solution for moving data into and out of Cassandra and manipulating data in it. This solution is free to use and supports both community and enterprise editions of Cassandra. It also connects to DataStax's big data platform that includes Cassandra and Hadoop distributions.

Apache Cassandra provides row-level atomicity of batch mutations, ensuring that multiple batched writes to the same row are applied by nodes atomically. However, prior to version 1.1, isolation of such updates was not guaranteed. In version 1.1 and later, row-level updates are made in isolation, guaranteeing that if you update both login and password for a user in the same update, no concurrent read may see only a partial update. These atomicity and isolation guarantees apply to columns written under the same physical row within the same column family and share the same partition key. For atomicity, the guarantee extends across column families (within the same keyspace). However, updates to different column families are not isolated. Internally, row-level atomicity is guaranteed mainly by the commit log and memtable structure of Cassandra's storage engine.

Cassandra, a distributed database system, initially followed a "schemaless" data model similar to Google's Bigtable paper. However, as systems grew and matured, the lack of schema became a pain point. Starting with version 0.7, Cassandra allowed users to define their data types, making it "schema-optional." The storage engine in Cassandra is sparse, allowing for easy addition of columns without reallocating space row by row. This flexibility is combined with the benefits of having a defined schema. In upcoming releases, CQL will support defining column families with compound primary keys, which can be useful for denormalizing data and faster queries.

The Cassandra File System (CFS) is an HDFS compatible filesystem designed to replace traditional Hadoop NameNode, Secondary NameNode and DataNode daemons. It simplifies operational overhead by removing single points of failure in the Hadoop NameNode and offers easy Hadoop integration for Cassandra users. CFS is modeled as a Keyspace with two Column Families in Cassandra: "inode" and "sblocks". The "inode" column family contains meta information about a file, while the "sblocks" column family stores the actual contents of the file. Meta information includes filename, parent path, user, group, permissions, filetype, and a list of block IDs that make up the file. CFS splits a block into sub-blocks since it relies on Thrift, which does not support streaming, to prevent overloading the node with large amounts of data at once. When a read comes in for a file or part of a file, CFS executes a custom Thrift call that returns either the specified sub-block data or, if the call was made on a node with the data locally, the file and offset information of the Cassandra SSTable file with the subblock. This approach cuts down network traffic between nodes by compressing and decompressing sub-blocks on the client side.

The blog post discusses how to install, setup, and manage/monitor Apache Cassandra on a Windows machine using DataStax Community Edition. It provides step-by-step instructions for downloading and installing the software, setting up the installation directory, handling services, launching DataStax OpsCenter, and accessing various utilities like CQL shell utility, Cassandra Command Line Interface (CLI), nodetool shell program, and managing services via Windows Services Control Panel. The post also mentions that DataStax introduced a free Windows installer for Apache Cassandra, making it easier to develop against the database on Windows.