ZebClient is designed as a unified solution, integrating the strengths of both parallel and distributed filesystems and combining that with a low-cost object storage for backend. By combining high-performance parallel file access with the scalability and low cost of object storage, ZebClient offers a versatile and efficient foundation for a wide range of demanding use cases. It is a data solution that addresses the needs of modern and extremely data-intensive applications.
Functions and Benefits Provided by the ZebClient Architecture
The ZebClient architecture provides ultra-high performance, scalability, cost-efficiency, and adaptability to modern data-intensive workloads. Its high-performance design, and resource optimization capabilities make ZebClient a robust data storage solution for distributed data-intense applications.
ZebClient provides the following functions:
Parallel File Access
ZebClient features a filesystem access for distributed simultaneous data retrieval and processing, which:
- Scales data-intensive workloads,
- Makes data available for distributed compute operations.
Distributed Access
ZebClient gives distributed access to data across multiple compute nodes, which:
- Provides accessibility to unified data across geographically dispersed locations,
- Delivers a POSIX-compliant filesystem for advanced workloads, where data consistency is secured and data is always available and updated across all nodes.
Object Storage Backend
The ZebClient architecture is based on object storage as the backend storage infrastructure, which:
- Enables seamless scalability by leveraging the object storage inherent architecture,
- Optimizes data storage, scaling its size dynamically from zero based on need,
- Secures low cost.
High Performance Design
ZebClient is designed to maximize data access performance, using techniques like:
- Storage tiering mechanisms placing frequently accessed data close to applications,
- Optimized data access and retrieval techniques such as striping,
- Option to add a local warm data tier to enhance data access performance to on-premise deployments.
Scalability and Elasticity
ZebClient supports horizontal scaling to accommodate a need for higher compute capacity. The solution also provides:
- Kubernetes-native deployments enabling dynamic resource allocation to handle varying workloads,
- Data storage scalability using cloud-based object storage to provide a filesystem unlimited in size.
Cost-Effective Resource Utilization
ZebClient maximizes the utilization of underlying resources and minimizes cost by providing:
- Optimal object data storage where the storage size is based on actual usage, avoiding the over-provisioning of traditional filesystems,
- A pay-as-you-go business model aligned with cost-effective cloud storage models.
High-Performance Data Access
ZebClient is designed to secure cost-efficient, high performance data access for a wide spectrum of workloads, which:
- Supports use of low-cost object storage for demanding applications such as analytics, AI and machine learning,
- Provides performance uplift to traditional file-based operations,
- Ensures compatibility for seamless integration across diverse environments and data formats.
Data Resilience
ZebClient uses the Zebware proprietary erasure code ZebEC alongside intelligent tiering algorithms which:
- Secures data integrity and availability from its redundancy mechanisms in all deployments where a warm tier is used,
- Immediately writes new data to persistent cold tier enabling data recovery in case of breakdowns or malicious attacks on the infrastructure.
Simplicity
ZebClient provides operational simplicity by:
- Presenting all data, including external S3 data in filesystem format, significantly improving visibility and reducing data administration efforts,
- Easy-to-handle deployments and updates of infrastructure including the Kubernetes-native design for optimal scale-out and scale-in according to current needs,
- Preventing vendor lock-in as ZebClient disaggregates compute from storage and enables easy portability of compute resources.