NoSQL databases exist since the late 1960s, but only since the beginning of the 21st-century databases, like ArangoDB, grew in popularity. The main reason is the rising popularity of social media platforms, but those are not the only use cases. In the following, we show a small example how modern NoSQL databases support the next generation of web applications.
What is ArangoDB?
ArangoDB is a native multi-model NoSQL database system. It was first released by the developer triAGENS GmbH under the name AvocadoDB in 2011. It has one database core and supports three data models: documents, graphs, and key/value. This multi-model approach offers high flexibility to master typical and highly specified web application tasks easily. ArangoDB’s database system runs on the query language AQL, which has similarities to SQL but differs by not using predefined sequences. However, despite all the differences, it should not be a problem with SQL experience to understand AQL. Due to its modern architecture, most of ArangoDB’s processes are automatized. This is especially useful for startup companies, which don’t have the necessary resources to maintain a database properly.
Why we need ArangoDB?
ArangoDB excels when used as a specialized database. The strength of a specialized database lies in its ability to provide exact search results from one or more subject areas. This is particularly advantageous when searching for very specific scientific topics. Cloud-based platforms such as iPlexus can greatly benefit from such a database. iPlexus offers a highly specialized and accurate search engine for life science topics including diseases, complicated protein structures, and drugs. The higher the precision of results, the more successful the product. ArangoDB’s runs on a C++ core, which provides one of the best benchmarking results of all database systems. A major strong point is, that it can run queries that traverse a search path of unknown length and find the best search results effectively. Furthermore, ArangoDB is freely scalable on one server system or a whole server cluster. It provides an easy-to-maintain and secure system that automatically replicates and secures data. In the event of server failure, ArangoDB provides an automatic failover process, which moves applications to standby
servers so that an all-time availability can be ensured. With ArangoDB you not only have one database for a variety of applications, you also have the possibility to perform ad-hoc queries of data, which are then stored in different models. In ArangoDB, you can use the same collection for a graph and for a document query without compromising performance. It is also possible to use API for complex graph transitions.
What are the different features and key differentiators of ArangoDB?
How are Innoplexus products powered by ArangoDB and what benefits we are getting with this engagement?
Our products kPlexus and iPlexus strongly rely on a fast, highly scalable and secure database system. This is due to our constantly growing data coverage in life science and potential future products in, for example, the financial sector. Unlike most other NoSQL database systems ArangoDB provides us with a one system solution. In the past, our products had a huge dependency on a variety of data stores and search engines including MongoDB, Elasticsearch, Neo4j & Redis. With ArangoDB we reduced data redundancy, maintenance time and benchmarking by integrating all those systems. One of the main advantages of ArangoDB is its performance in the areas of single read & write, SPFA and NNS. Only in for example area of data aggregation and memory usage some of the competitor databases can deliver better results. So what consequences does this have? In essence, all these factors are an indication of how efficient a database can operate. Data aggregation combines different datasets into one information, while single read and write set the pace of how fast a data set can be processed. Shortest Path Faster Algorithm (SPFA) scans the database for the shortest connection between a given point of the data network. Nearest Neighbour Search (NNS) on the other hand, aims to find data records with a similar meaning. Since we wanted to choose the best database, we were interested in overall performance. For this reason we chose ArangoDB as our new database system and improved performance, reduced maintenance and have laid the right foundations for unrestricted growth.
ArangoDB powered kPlexus: Identifying the key KOL amongst the KOL networks in a therapeutic area.