November 2019 Summaries

This summary provides an overview of the process to quickly set up a minimal SingleStore cluster using Docker Desktop on a laptop. The steps involve getting a free SingleStore license, installing Docker Desktop, creating a `docker-compose.yaml` file, starting the SingleStore cluster through Docker, and browsing to SingleStore Studio. The setup allows for easy experimentation with SingleStore capabilities, prototyping, and testing without requiring significant hardware resources. Once the setup is complete, users can connect to the cluster using SingleStore Studio, execute queries, and explore the features of the database engine. The summary highlights the convenience of this setup for "kick the tires" scenarios and encourages readers to get a copy of SingleStore for free to start building great things.

The company was struggling with its analytics platform, which had several issues such as stale data, loss of information, clumsy processes, operational complexity, and not being future-ready. They wanted a simpler architecture that could deliver real-time operational analytics, machine learning, and AI capabilities. After investigating several options, they chose Kafka and SingleStore to replace their current platform. The new solution aims to simplify the data processing core, reduce costs, and increase performance. It will use Airflow for orchestration, Kafka as a pipeline capability, and SingleStore as the database. The company is excited about the potential of SingleStore, which they believe will solve many of their problems and provide a scalable platform that can handle large amounts of data and complex queries.

SingleStore is presenting a workshop on November 12th in New York City at the AWS office, focusing on enabling real-time data-driven insights using Amazon Sagemaker and SingleStore, a highly performant cloud-native database. The workshop will cover how to deploy machine learning models into production, leveraging SingleStore's capabilities such as streaming data analytics, operational analytics workloads, and SQL-compatible tools. SingleStore integrates with popular machine learning tools like TensorFlow, scikit-learn, and R, allowing for rapid model deployment and high-speed scoring. The workshop will also demonstrate how to jointly deploy SingleStore real-time streaming ingest pipelines with Sagemaker inference ML endpoints, enabling organizations to reduce time to value and operational costs while improving the customer experience.

Singlestore Helios, a database-as-a-service offering, was created using Kubernetes, which provides stateful application support. This allows for easier deployment and maintenance of Singlestore databases across multiple cloud providers, including AWS, GCP, and Azure. The platform aims to empower customers to deploy their databases on the infrastructure of their choice, improving portability and ease of management. To achieve this, a custom Kubernetes Operator was developed, which interacts with the SingleStore database via an engine interface called memsqlctl. This enables features like online upgrades, declarative configuration, recovering from failure, scalability, and auto-scaling, making it easier for customers to manage their databases. The platform's flexibility and scalability will enable new features and capabilities to be built on top of it, improving operational simplicity and customer experience.