On-premise, Cloud-based, or hybrid
Deployable where needed, whether onsite, on public or private clouds, or within a hybrid environment.
The Sinequa platform is provided on-premises and fully compliant with cloud stacks that Microsoft, AWS, and Google propose. Smart implementations are often hybrid to
- Simultaneously cope with complex internal data sources and associated security rules; and
- Leverage the flexibility and lower TCO of public cloud service offerings.
Sinequa dedicated components allow you to collect data sources locally before securely forwarding the minimum required data to a remotely located grid architecture that can be hosted on a cloud or in a different network or geography.
Sinequa integrated within Azure streamlines operations and makes platform deployment fast and easy.
Streamlined cloud operations for accelerated deployments, including:
- Availability on Azure Marketplace
- Pre-set templates
- Scripts to build Sinequa's images
- Sample scripts (PowerShell, Terraform, ARN, etc.)
- Dynamic indexing
- Embedded security, thanks to an integration with Azure key vault
- Integration with Microsoft Teams
Sinequa's platform capabilities scale every which way to ensure a performant, high availability experience for all employees.
Sinequa offers a unique grid architecture to handle very large data sets that must be ingested in bulk.
- Modular and distributed: more than one server needed for volumes above 100 million documents or/and high-availability requirements
- Scalable: ability to host tens of billions of index entries
- Highly available: dynamic index replication, load balancing, dynamic clustering, etc.
- Performant: hundreds of queries per second and search node
Each node is configured for one of the following roles:
- Application server (Sinequa WebApp)
- Search server (Sinequa Engine)
- Indexing server (Sinequa Indexer)
- Sinequa smart connectors and converters
This unique architecture strikes a balance between various forms of scalability and uncompromised performance. The grid approach enables horizontal scaling such that nodes can be easily added to the grid - for example, additional servers dedicated to indexing. The approach also enables vertical scaling - for example, leveraging multi-core processors for parallel processing. As the solution scales in either or both directions, performance is maintained as the Engine and Indexers are independent processes, meaning that indexing can run simultaneously as querying without penalizing response time.
Scalability, high availability & load balancing
Sinequa provides almost instantaneous results, generally under a second, and can manage up to hundreds of queries per second on a single dedicated server. Sinequa's platform is highly scalable and can easily be expanded to accommodate more content from existing or new sources.
All components of the Sinequa architecture can be distributed and spread across the enterprise network and operate efficiently on physical and virtual nodes.
Sinequa easily supports a high-availability architecture that replicates indexes across multiple geographically distributed nodes and shares the load of the queries sent by users across several engines.