Operational Data Lake
Operational Data Lake（ODL）can be considered the next-generation iteration of operational data store（ODS）, with the latter often built on top of proprietary commercial databases (like Oracle). ODL enables certain types of workloads that are often hard to support on either traditional OLTP or OLAP databases. For example, running complex queries or reports on OLTP system or allowing real-time updates on OLAP systems. New types of ODL should enable near realtime analytics and reporting to be done close to the data source while performing high throughput data ingestion and transformation (stream or batch).
Compared to the scale-up paradigm of ODS systems, ODL can horizontally scale with commodity hardware, offering 20 times better TCO (total cost of ownership). Besides, ODL can store and process structured, semi-structured, as well as unstructured data, making it part of a bigger big data system environment.
Complex and inefficient ETL process is usually the hidden liability of corporate IT departments. For Big Data platforms, ETL can quickly becomes the bottleneck for downstream data analysis and applications. Different from using heavy proprietary database systems (like Oracle or Teradata), customers can leverage LongDB data platform's capability to reduce ETL process delay from days or hours to minutes or seconds. This is possible due to LongDB's innovative architecture and the seamless integration with streaming and batch systems (Kafka, Flink, Spark, Hive, Sqoop, etc.)
Replacing Old Data Warehouse
There is a growing market for enterprise data warehouse, thanks for growing data volume and analytical needs. However, the EDW market used to be dominated by vendors like Oracle, Teradata, and IBM. These corporations built EDWs using their proprietary technologies which typically require expensive purpose-built hardware appliances. However, with growing business, data volume grows along with higher demand on analytical requirements and faster response time. Their old architectures cannot keep the growth without incurring inhibitive cost increases. Some new data-driven applications also have grown out of the traditional relational database use cases, with AI/ML technologies added to uncover values of data. LongDB platform combines the best of both worlds, supporting both the traditional database capabilities and new types of AI/ML use cases with ever-evolving open source ecosystem components.
Powering IoT Applications
Most big data systems for IoT applications have been built using open source technology in a "DIY" fashion because most of these applicatios are relatively new and none of the traditional system could handle well. Open source modules often lack the desired maturity or functionality (although improved greatly over time), which lead to the introduction of additional components and in-house customizations. Glue code is used to make all pieces work together. There is no holistic data integrity and quality control, no strict user authorization and governance, which leads to higher and higher development and maintenance costs over time.
LongDB's innovative data platform perfectly solves the aforementioned problems. It provides linear scalability on data storage and processing, which is often requires by IoT systems fast growing data volume and analytics requirements. The platform combines realtime data ingestion as well as analytics capabilities without the need for customers to have a deep pocket. The versatile platform can also support different kinds of data (structured, semi-structured, and unstructured) to enable users to build limitless applications to utilize the data.