March 2015 Summaries

Graph databases are designed to scale out across a distributed cluster, with TitanDB utilizing Cassandra's distributed storage engine to achieve this advantage. In contrast, Neo4j scales up and relies on a master/slave architecture, requiring more powerful machines for scaling. The choice of approach depends on the tradeoffs between performance, reliability, and scalability. Some databases, like OrientDB, offer hybrid document-graph engines that blend SQL with graph manipulation capabilities. These differences highlight the varying design philosophies and technical approaches of graph databases in a distributed platform.

Neo4j is a highly regarded graph database used by thousands of organizations, including many Global 2000 companies, in mission-critical production applications. It boasts the largest and most active community of graph database enthusiasts and has been recognized for its operational database magic quadrant standing by Gartner. Neo4j delivers exceptional performance, scalability, and reliability, making it an ideal choice for enterprise and startup projects alike, with features such as native graph storage, ACID compliance, and a user-friendly interface that supports various programming languages and development tools. With its staggering loading speed, intuitive data modeling capabilities, and superb value proposition, Neo4j is well-suited to help organizations unlock the full potential of their data and applications.

The new Neo4j 2.2 release includes several exciting features that make it easier and more intuitive for users to get started with the graph database. The `:play` command allows users to input data into Neo4j in a gentle and easy-to-use manner, while also providing an interactive experience that helps users learn and understand the technology. The query pane has been redesigned with several enhancements, including the ability to cancel running queries, export data to various formats, and display auto-completion options. Additionally, new browser commands such as `:style` and `:config` provide fine-grained control over the visualization and behavior of the Neo4j Browser application. The Browser also includes a feature to create composite captions on visualizations, which allows for more complex and dynamic node labels. Finally, the query planner now displays query plans visually, making it easier for users to understand and tune their queries.

Neo4j has released version 2.2, a significant update that offers substantial improvements in performance, scalability, and developer productivity. The new release boasts massive write throughput, up to 100 times higher than previous versions, allowing for faster data loading and concurrent transactions. Additionally, Neo4j 2.2 introduces a cost-based optimizer for Cypher queries, resulting in up to 100 times faster query performance. The update also includes a brand-new in-memory page cache designed to deliver extreme performance under highly concurrent workloads. Furthermore, the release provides numerous developer-friendly features, such as quick start guides, code examples, and improved visualization tools. Overall, Neo4j 2.2 represents a substantial step forward for graph databases, enabling organizations to derive maximum value from their data relationships.

The new version of Spring Data Neo4j is a server-first, Cypher-centric project that aims to exceed the performance of existing SDN, targeting Neo4j in traditional deployments where the database can scale independently of application servers. The new implementation introduces innovations such as non-reflection based classpath scanning for faster startup times and variable-depth persistence to fine-tune requests according to graph characteristics. Existing SDN users should be aware that there are breaking changes, but the APIs have been preserved with minor differences. A new example application and a university administration system using Angular.js and Spring-Boot are available on GitHub, allowing users to explore the new version and provide feedback through the project's issue tracker or email. The development team is grateful for contributions from Neo4j engineers, GraphAware team, Mark Angrish, and Oliver Gierke, who helped significantly with the Spring integration.

The Neo4j 2.2 Meta-graph query has been improved to better visualize relationships in the browser. Originally, the query created nodes for every relationship and used those nodes to represent both node-labels and relationship-types, which was confusing. However, with the new query, a separate "name" property is added to the `META_RELATIONSHIP` relationship type, allowing users to select this property in the browser to create a more accurate visualization of relationships. This change enables a cleaner representation of nodes as labels and relationships as types, making it easier to understand complex connections between data entities.

The query used in the training sessions is intended to find actors and directors who have worked together on movies, but a common typo can result in no results. The "EXPLAIN" keyword has been introduced in Neo4j 2.2 to help diagnose issues with queries like this one. By using "EXPLAIN", users can see what the query planner thinks the query would do without actually running it, including an estimated number of rows that need to be processed. This can help identify typos or other problems with the query.

The author, Rik, explains the process of explaining a graph model in an existing Neo4j database. He shares a query that generates a table showing the labels and types of relationships between nodes in the database. To visualize the model graphically, he tweaks this query to create a META-graph by merging nodes for each label and relationship, using MERGE statements. The resulting diagram is readable and easy to understand, making it useful for visualizing existing graph models and databases.

The text appears to be a summary of the QCON London 2015 event, where several speakers shared their expertise on Neo4j and graph databases. Mark Needham gave a talk titled "Graphs R Cool" and was also involved in a Q&A session. Other speakers discussed topics such as graph visualization tools, data provenance, and the application of graph databases to various domains like the Internet of Things. The event also featured a meetup group analysis using Neo4j, which highlights its potential for modeling complex relationships and structures.