VAST DataSpace

Say Goodbye to Data Gravity

Meet the revolutionary VAST DataSpace.​ A continuation of our first-principles thinking, the DataSpace​ provides consistency and fast access to your data, from edge to cloud.​

Read the White Paper
Features
Resources

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, across 100’s of locations.

Breaking the tradeoffs between performance and consistency.

VAST had to completely rethink how lock management was done across hybrid environments.  Existing implementations were optimized for read-centric workloads and fall flat with transactional workloads.​

VAST DataSpace decentralizes and moves the lock management down to the element-level (file, object, table) and allows for the lock to be held by the VAST cluster doing the transaction.​

Reads can achieve peak performance while writes maintain consistency because before writes are executed by a VAST cluster, it will ensure that all globally cached copies of that element are removed and any references will be directed to the locking cluster until the transaction is complete.​

Access your data anywhere — without all the tradeoffs.

Snapshot copies of data at the volume level lack data coherency. This can lead to potential data inconsistencies and issues when trying to restore or access specific points in time. On the other hand, centralized lock management with directory-level locks can cause write bottlenecks because all write requests must pass through a central point. With the VAST DataSpace, the system delivers fine-grained de-centralized locks, delivering local write performance. By distributing the lock management across different parts of the system, it improves concurrency and enhances overall system performance, ensuring data integrity and efficiency in handling read and write operations.

Flexible, path-based synchronization.

VAST DataSpace allows sites to subscribe only to what they want, making it a more efficient solution out the gate. Often times there is no need to synchronize the entire dataset across remote sites.

Each site can subscribe to any path and access the data through the protocol of choice: file, object, and table. This allows for flexibility in supporting 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.

Scheduled Mode

In Scheduled Mode, VAST DataSpace utilizes custom scripts or popular job schedulers (Run AI, Slurm, Kubernetes) to anticipate data needs, balancing efficiency and performance for remote sites.

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.

images

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.

images

Your Data, Your Choice of Location

VAST Data Platform has been available through manufacturing and systems partners to deploy enterprise-grade infrastructure in your core and edge datacenters, in both ruggedized and non-ruggedized environments.

Now the VAST Data Platform is available through various Cloud Service Providers such as AWS, Azure, and GCP.

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.

images
Features

Designed for AI, the Enterprise, and Beyond​

Replication

Replication is how the VAST DataSpace maintains copies of datasets across multiple locations. The VAST Data Platform provides multiple replication methods so customers can provide each dataset the level of protection and recovery time objective it deserves.​​

Snap-to-Object​

Snap-to-Object leverages VAST’s low-cost snapshot mechanisms to create a backup copy of a folder’s data on an S3 compatible datastore.

Asynchronous Replication for Data Durability​

The VAST Data Platform periodically replicates the changes between snapshots across VAST clusters over one-to-one, one-to-many and many-to-one topologies.

Synchronous Replication for RPO Zero​​

Synchronous replication creates an active-active cluster, supporting strictly consistent writes across a pair of VAST clusters for applications that require zero downtime or data loss.

Replication

Replication is how the VAST DataSpace maintains copies of datasets across multiple locations. The VAST Data Platform provides multiple replication methods so customers can provide each dataset the level of protection and recovery time objective it deserves.​​

Snap-to-Object​

Snap-to-Object leverages VAST’s low-cost snapshot mechanisms to create a backup copy of a folder’s data on an S3 compatible datastore.

Asynchronous Replication for Data Durability​

The VAST Data Platform periodically replicates the changes between snapshots across VAST clusters over one-to-one, one-to-many and many-to-one topologies.

Synchronous Replication for RPO Zero​​

Synchronous replication creates an active-active cluster, supporting strictly consistent writes across a pair of VAST clusters for applications that require zero downtime or data loss.

Global Access​

Global Access allows VAST customers to make folders and buckets available to servers and users on multiple VAST clusters as if they were local to every cluster in the DataSpace. Flexible access models range from independent clones to a strictly consistent global namespace.

Clone Snapshots Across The DataSpace​

Global Clones create content-independent clones from one VAST cluster to another across the DataSpace.​​

Global Access Creates a Global Namespace​

VAST Global Access provides read/write access to global folders across VAST clusters at the edge, core, and cloud, creating a global namespace.

Read and Write Leases Are Smarter Than Locks

Granular read and write leases provide full read/write access with all-flash performance across multiple locations, all while ensuring strict consistency.

Global Access​

Global Access allows VAST customers to make folders and buckets available to servers and users on multiple VAST clusters as if they were local to every cluster in the DataSpace. Flexible access models range from independent clones to a strictly consistent global namespace.

Clone Snapshots Across The DataSpace​

Global Clones create content-independent clones from one VAST cluster to another across the DataSpace.​​

Global Access Creates a Global Namespace​

VAST Global Access provides read/write access to global folders across VAST clusters at the edge, core, and cloud, creating a global namespace.

Read and Write Leases Are Smarter Than Locks

Granular read and write leases provide full read/write access with all-flash performance across multiple locations, all while ensuring strict consistency.

VAST on Cloud

The VAST Data Platform extends to the cloud not just as the backend behind the biggest AI cloud providers but also as VAST clusters on AWS, GCP and Azure. These VAST on Cloud clusters allow cloud compute resources access to the data in your VAST DataSpace.​​

VAST on Cloud Instances

VAST on Cloud instances extend the VAST DataSpace into the public cloud, enabling hybrid workloads

VAST Cloud Clusters

VAST on Cloud clusters provide the functionality and scalability of the VAST Data Platform on the public cloud.

Archive Clusters

VAST on Cloud Archive Clusters extend the VAST DataSpace to cloud object storage for cold archives. Archive Clusters then contribute their contents to a global namespace, which provides transparent, high-performance access to the archived data.

VAST on Cloud

The VAST Data Platform extends to the cloud not just as the backend behind the biggest AI cloud providers but also as VAST clusters on AWS, GCP and Azure. These VAST on Cloud clusters allow cloud compute resources access to the data in your VAST DataSpace.​​

VAST on Cloud Instances

VAST on Cloud instances extend the VAST DataSpace into the public cloud, enabling hybrid workloads

VAST Cloud Clusters

VAST on Cloud clusters provide the functionality and scalability of the VAST Data Platform on the public cloud.

Archive Clusters

VAST on Cloud Archive Clusters extend the VAST DataSpace to cloud object storage for cold archives. Archive Clusters then contribute their contents to a global namespace, which provides transparent, high-performance access to the archived data.