September 2011 Summaries

Cassandra 1.0 introduces several improvements to its storage engine, enhancing memory and disk management for better performance. Key updates include the addition of an off-heap row cache for reduced JVM heap requirements, self-tuning memtable sizes for optimal balance between write speed, compaction overhead, and memory use, a global commitlog_total_space_in_mb setting to manage data replay on startup, and explicit reference counting for efficient obsolete data file cleanup post-compaction. These enhancements address common pain points for Cassandra administration and improve overall performance.

Cassandra 1.0 introduces support for data compression on a per-ColumnFamily basis, which maximizes storage capacity by reducing data volume on disk. Compression also reduces disk I/O and improves both read and write performance. It is best suited for ColumnFamilies with many rows having the same columns or many columns in common. Enabling compression can be done during column family creation or update, and existing SSTables are compressed during normal Cassandra compaction process.

Cassandra, a distributed database management system, has introduced a new tool called sstableloader to improve bulk loading of data. This tool streams sstable data files to live clusters and only transfers relevant parts of the data to each node, conforming to the replication strategy of the cluster. The primary use cases for this tool include transferring data from one test cluster to another multi-node cluster and bulk-loading external data that is not in sstable form. To use sstableloader, users need to configure a cassandra.yaml file with correct settings and ensure that the schema for column families is defined beforehand. Additionally, users can create relevant sstables from CSV files using the SSTableSimpleUnsortedWriter class introduced in Cassandra 0.8.2.