The VAST DataSpace has broken the fundamental tradeoff between performance and consistency to make it possible to access, transact, and protect data from edge to cloud.
The VAST DataSpace is a global namespace that permits every location to store, retrieve, and process data from any location with high performance while enforcing strict consistency across every access point.
Given that data cannot travel faster than the speed of light, the globe that a global namespace can effectively serve with high performance and strict consistency is practically limited to a data center neighborhood, say, all locations in northern Virginia.
​
The VAST DataSpace solves that problem by joining VAST clusters into a constellation forming a Global Path where folders and hosts in multiple, geographically dispersed locations can be simultaneously presented to users, delivering local performance for read-write access.
​
​Users can read or write to this Global Path as if it were just a folder in the local namespace.
Ensuring Strict Consistency – Rethinking Distributed Lock Management
VAST has completely rethought how lock management is done across hybrid environments in order to ensure strict consistency for transactional workloads.
The VAST DataSpace decentralizes and moves lock management down to the individual element-level (file, object, table) and allows for the lock to be held by the VAST cluster doing the transaction. With fine-grained de-centralized locks implemented with innovative read and write leasing mechanisms, the Dataspace delivers local write performance.
Your Data, Your Choice of Location
VAST Data Platform has been available through systems partners to deploy enterprise grade infrastructure in your core and edge datacenters.
Now the VAST Data Platform is available through various Cloud Service Providers such as AWS, Azure and GCP.
Reads are up to date and maintain peak performance
Before writes are executed the VAST cluster will ensure that all globally cached copies of an element are removed and any references will be directed to the locking cluster until the transaction is complete.
By distributing the lock management across different parts of the system, the Dataspace improves concurrency and enhances overall system performance, ensuring data integrity and performance in handling read and write operations.
With VAST DataSpace, customers are able to seamlessly move their data to where it needs to be without having to use third party software or maintain complex data orchestration environments.
Your data is where it needs to be, when it needs to be.
Flexible, Path-Based Synchronization and Consistency Modes
Oftentimes there is no need to synchronize the entire dataset across remote sites. VAST DataSpace allows sites to subscribe only to what they want. Each site can subscribe to any path and access the data through the protocol of choice: file, object, and table in support of all the various applications across your organization. Each path can have its own set of synchronization policies and consistency modes allowing you to support the diverse set of requirements of your organization.
Lazy Mode
​
VAST DataSpace’s Lazy Mode intelligently syncs data based on access predictions, maximizing bandwidth utilization. It’s storage-efficient but has a higher time-to-first byte.
Lazy Mode
​
VAST DataSpace’s Lazy Mode intelligently syncs data based on access predictions, maximizing bandwidth utilization. It’s storage-efficient but has a higher time-to-first byte.
Eager Mode
​
VAST DataSpace will ensure data is immediately cached in remote sites to ensure instant data access is achieved. Although this is less storage efficient, it optimizes for performance.
Read/Write Paths Consistency Mode
​
In Read/Write Mode, all sites in the DataSpace will be able to read and write with consistency, while providing a global namespace with both read and write performance you can expect from a local storage system.
Global Clones Consistency Mode
​
Writeable Clones are great for use cases when you want a point-in-time copy of particular path of your namespace, but you don’t want the updates to propagate back to the global namespace.
PDF - The VAST Data Platform White Paper
This comprehensive White Paper describes The VAST Data Platform in detail.
Pages 69 through 73 describe the VAST DataSpace, including the read and write leasing mechanisms employed to ensure strict data consistency for both writes and reads.
The VAST Data Platform is a breakthrough approach to data-intensive computing that serves as the comprehensive software infrastructure required to capture, catalog, refine, enrich, and preserve data through real-time deep data analysis and deep learning.
It is designed to provide seamless and universal data access and computing from edge-to-cloud, all from a platform that is designed for enterprises and cloud service providers to deploy on the infrastructure of their choosing.
Summary of Features
Designed for AI, the Enterprise, and Beyond
Global Namespace
​
VAST DataSpace enables any or all of your data to be accessible anywhere you need it to be, without having to use third party software or complex data orchestration logic. Whether it’s in your core or edge datacenters, or in your cloud service provider, VAST will ensure your data will be where it needs to be.
Global Data Reduction for All-Flash at Archive Price Points
​
VAST’s Similarity-Based data reduction combines the global nature of deduplication with the fine granularity of compression across your entire global namespace. Combined with low-cost hyper-scale flash, VAST delivers a data platform with Tier-1 All-Flash performance that sells for a price point equivalent to archive storage.
Massive Scale
​
VAST clusters can be built to support well over an exabyte of data capacity. Today, several customers run clusters over 100PB in size.
Decentralized Lock Management
​
Locks are distributed to each node that needs it and are established at a very granular level on elements (files, objects, tables). This provides high transactional performance without having to check with a central repository.
Fine-grained Lease Revocation
​
Before writes are performed, VAST DataSpace will ensure all cached copies of that element are removed and reference the locking site until the write is completed. This ensures high transactional performance while maintaining write consistency.
Path-based Synchronization
​
Any number of sites can subscribe to any paths from any other site ensuring data is available where it needs to be, without having to unnecessarily synchronize unnecessary data.
Flexible Sync Policies
​
VAST DataSpace offers Lazy, Eager, and Scheduled Caching to allow users to optimize for storage efficiency, time-to-first byte, or optimize for scheduled jobs.
Flexible Consistency Modes
​
Read/Write modes allow for remote sites to simultaneously write to the same namespace while ensuring write consistency. Writeable clones allow for a point-in-time version of a path to be written to on a remote site without propagating those changes back to the global namespace.