StarTree Cloud provides managed hosting for Apache Pinot on all major cloud platforms, including AWS, GCP, and Azure. StarTree Cloud offers public SaaS (software-as-a-service) and private BYOC deployment options. For more information about our public or private SaaS offering, schedule a demo with our StarTree solutions architect team. Or, get started with a free 30-day trial.
StarTree Cloud lets you provision Apache Pinot clusters of different sizes, ingest data from real-time and batch data sources, and run analytics workloads with ultra-low latency. StarTree manages the underlying infrastructure for you, provides insights from a diverse set of data, and helps you make informed business decisions.
What are the use-cases for StarTree Cloud?
- User facing analytics
- Business metrics
- Anomaly detection
- Root cause analysis
- Log analytics
- Cohort analytics
- Ad hoc exploration
|Query throughput (queries/sec)||Query latency (p95th)||Consistency/accuracy||Query flexibility|
|User facing analytics||Very high: 10k-100k||10ms-100ms||Best effort||Low|
|Personalization||Very high: 10k-100k||10ms-100ms||Best effort||Low|
|Anomaly detection||Moderate: 100s-1k||10ms-100ms||Best effort||Low|
|Root cause analysis||High: 100s-10k||10ms-100ms||Best effort||Low|
|Visualization/dashboarding||Moderate: 100s-1k||100ms||Best effort||Medium|
|Ad hoc analytics||Low||Single-digit seconds||Best effort||High|
|Log analytics/text search||Moderate: 100s-1k||Sub-second||Best effort||Medium|
Key features of StarTree Cloud
Managed Apache Pinot
- Query Latency and Speed: Computes on the fly and is very fast due to the indexing strategies, partitioning/data layout, and bloom filters. StarTree supports partial pre-aggregated values to provide very low latency, real-time analytics on the data.
- Millisecond level latencies for most OLAP queries
- Allows to achieve a hard upper bound for query latencies for a given use-case
- Low latency with high throughput
- Data Mutability: Designed with low cost to serve to answer OLAP queries and low latency on immutable data and mutable data(Upsert Support).
- Indexing: The following indexing techniques are supported:
- Inverted index
- Sorted index
- Range index
- JSON index
- Text index
- Geospatial index
- Star-tree index
- Throughput: Purpose-built for supporting very high throughput for the analytical workload. Can support 10000+ QPS in a single cluster.
- Cost to serve: Low columnar storage provides excellent compression leading to lower storage and in-memory footprint.
- Operational/Production Readiness: Built to be multi-tenant. Provides an easy way to scale a cluster up or down, replace nodes, and reshuffle data.
- Advanced query features (joins): Limited support-dim join supported and an early version of fact-fact distributed shuffle join is also available.
- Integration with existing data eco-systems: Integrates well with the rest of the data ecosystem with support for backfills.
- Monitoring with alerts sent to StarTree SRE team: StarTree provides service-level agreement (SLA) guarantees. To assist in meeting those guarantees the StarTree-managed Pinot environments have a comprehensive set of metrics being monitored. In addition, Prometheus alerts are configured. It is not expected that customers will receive these alerts, but instead that the StarTree site reliability engineering team receives and fixes all alerts. We're committed to notifying affected customers of both alerts and alert resolutions.
StarTree Data Manager
The StarTree Data Manager is a UI that makes it easy to onboard data into a StarTree Cloud.
To open Data Manager, go to StarTree Cloud, and click Data Manager.
For more information, see the following: