November 2013 Summaries

Titan is a distributed graph database designed for handling large-scale graphs with billions of edges. It supports both single machine and highly available cluster configurations, making it suitable for various applications ranging from small boutique to massive Big Graph Data applications. The latest version, Titan 0.4.1+, introduces significant improvements in memory management and caching mechanisms, leading to faster traversals and better performance. This update enables seamless scaling of boutique graph applications as they grow into larger-scale applications requiring distributed processing power.

Apache Cassandra offers various features to secure data, including support for enterprise-grade authentication using Kerberos through DataStax Enterprise (DSE). The DataStax Java driver can be used with both DSE and Apache C* for secure access. A DSE-specific version of the java driver is available for use with secure DSE clusters, which includes additional authentication classes such as com.datastax.driver.core.sasl.DseAuthProvider. This driver supports Kerberos options set using a standard JAAS Login configuration file. Additionally, Apache Cassandra supports client to server encryption using SSL, which can be enabled programmatically or by configuring system properties.

The company has open-source its C++ Driver for Apache Cassandra, continuing their initiative to provide simple and modern APIs for Cassandra in major programming languages. This new driver is designed for CQL and the CQL Binary Protocol, targeting Cassandra 1.2+. It's based on Matt Stump's work on libcql, with further development planned within DataStax. The driver currently supports all major features found in other drivers but will be tested heavily before release. A beta version is expected soon.

Apache Cassandra® introduced internal metrics exposure using the Metrics library starting from version 1.1, with a significant expansion in version 1.2 and beyond. These metrics cover cache size, hit rates, client request latency, thread pool status, per column family statistics, and other operational measurements. Since Cassandra 2.0.2, pluggable Metrics reporter support is built-in, allowing users to send these metrics to systems like Graphite or Ganglia for data storage and graphing. The metrics can be customized by including or excluding specific groups of metrics based on the cluster's requirements. These metrics are useful for monitoring and troubleshooting performance issues in a Cassandra cluster.

The latest minor releases of Cassandra 2.0.0 contain numerous bug fixes as well as new features and enhancements. Some highlights include improvements to handle larger columns, allowing SELECT DISTINCT on partition keys, optimizing cleanup operations, logging statistics for repair operations, introducing the LOCAL_ONE consistency level, saving compaction history results, adjusting default write request timeout settings, and prioritizing compacting frequently read data files. These updates aim to improve availability, latency, and overall performance of Cassandra systems.

The text discusses how marketers create artificial distinctions between products to grab consumer attention, using Diamond Shreddies as an example. It then delves into the technology sector, specifically focusing on graph databases. The author argues that there is no such thing as a native graph database and explains why it's difficult to achieve perfect mapping of graphs in sequential memory layers. They also discuss how different graph databases handle this challenge and provide an overview of Titan's approach. The text concludes by encouraging readers to ask vendors about their graph embedding strategies when considering "native" graph databases.