StarTree addresses enterprise requirements that have been long solved in batch processing but present unique complexities for real-time analytics.
StarTree has unveiled a set of new innovations designed to equip organizations to efficiently handle evolving data structures, enhance query performance, and streamline user access management, ensuring faster and more reliable real-time analytics at scale.
As organizations shift towards real-time analytics, they face the dual challenge of adapting to an environment where everything is expected instantly and at an unprecedented scale. The rapid evolution of table sizes, the growing number of tables, and soaring ingestion and query rates amplify the complexity of managing dynamic data structures. Apache Pinot demonstrates exceptional scale, powering real-time workloads at LinkedIn with upwards of 650,000 queries per second, Stripe managing approximately 1 PB of data, and Uber achieving a 99th percentile query latency of just 100 milliseconds.
Unlike batch systems, which benefit from stable, periodic data loads, real-time analytics requires solutions that maintain performance, security, and reliability amidst ever-changing conditions and at unprecedented scale.
Changes in streaming pipelines – such as schema shifts or data gaps – require immediate resolution and optimized queries are essential for sustaining throughput and responsiveness in customer-facing applications.
StarTree Cloud’s new capabilities address these unique needs, enabling organizations to manage real-time data efficiently while maintaining stringent performance and security standards.
New innovations in StarTree Cloud, include:
Pauseless Ingestion: Pauseless ingestion underscores StarTree’s dedication to data freshness at scale, ensuring that every second counts. As organizations ingest tens of millions of messages per second, even brief delays can impact data accuracy and decision-making. StarTree Cloud maintains continuous data flow during segment building and upload phases, enabling businesses to deliver real-time, reliable insights at scale.
Performance Manager: Designed to help teams scale up quickly and minimize time to value when using Apache Pinot. With an intuitive, machine-learning-powered interface, it simplifies the process of optimizing query performance. New users often face challenges navigating the extensive range of indexing technologies available. Performance Manager addresses this by analyzing query structures and metrics to recommend enhancements—such as indexes, bloom filters, derived columns, or star-tree indexes. Users can apply these optimizations with a click, achieving immediate performance gains. This automation not only accelerates onboarding and efficiency but also maximizes cluster throughput, reducing manual effort and boosting overall system performance.
Schema evolution: In real-time databases, where data flows continuously, schema evolution in StarTree Cloud allows the system to accommodate new fields, indexes, altered data types,or other structural modifications without disrupting operations. This capability is essential for maintaining data consistency and ensuring that applications relying on the database continue to function smoothly despite the evolving nature of the input data.
Data Backfill: This feature addresses incorrect or missing data by enabling users to seamlessly reload data from past events, filling any gaps in data flows. When data fails to load or stream correctly, backfill allows teams to go back and retrieve the incorrect or missing information, preserving consistency across datasets. This capability is particularly valuable in real-time analytics, where continuous data integrity is essential. By automating the backfill process, organizations can easily maintain accurate, up-to-date information.
Role-Based Access Control (RBAC) Management: Enhances security and simplifies user administration for real-time data analytics. This feature allows organizations to easily assign and control user access based on roles, ensuring secure, efficient access to sensitive data, even when that data is ingested and analyzed by that role in a sub-second window.
Kishore Gopalakrishna, Co-founder and CEO, StarTree, said: “StarTree’s new features bridge a crucial gap in real-time analytics, enabling organizations to scale with the reliability and control typically seen in batch systems. By tackling critical challenges in data management and security, StarTree helps organizations better leverage real-time insights, enhancing their ability to scale operations efficiently.”
Paul Nashawaty, Principal Analyst, AppDev, theCUBE Research, said: “Real-time analytics presents a distinct challenge compared to batch-based analytics, as it requires instant adaptation to data changes while maintaining stringent performance demands.
“Industry research found that 75% of organizations struggle with the complexity and latency associated with traditional real-time analytics solutions. StarTree is bringing essential improvements to real-time analytical databases, paving the way for broader adoption of real-time analytics across many industries. It enables businesses to gain real-time insights from their data and make faster, more informed decisions.”
