VAST Data Release Train Is Packed With New Capabilities

VAST Data has announced software release 4.6 + 4.7, calling it the largest software release in its history. It is so big Jeff Denworth, VAST Data CMO, compiled a blog that outlined the capabilities and indicated there would be more to come because it couldn’t be done in a single blog post.

VAST Data has grown the R&D team to about 300 engineers to rival large enterprise software and cloud infrastructure teams. This growth allows VAST Data to invent and develop VAST OS simultaneously in multiple directions and dimensions. If you want some background, we did an in-depth piece on VAST Data Ceres in March 2022. Brian also has a podcast with Jeff if listening is more your speed.

Spotlight on the VAST Data Catalog

Organizations have followed the rule that database infrastructure should be independently built alongside content stores to make sense of unstructured and semi-structured data. This level of complexity makes it virtually impossible to get complete information from any given data store. IT is accustomed to the idea that structured, unstructured, and semi-structured datastore should be distinct because a single system had yet to be designed to achieve accurate data synthesis.

One of the benefits of VAST Data architecture is the ability to scale up to exabytes of capacity across billions of files and objects. Managing all that data can be challenging. To address the management challenge, VAST created a built-in metadata index called the VAST Catalog that makes searching and locating data easy in a fraction of the time traditional methods take. With this technology, users and applications can start treating the Filesystem like a Database, enabling next-generation AI and ML applications to use it as a self-referential feature store.

The Element Store is the center of VAST Data Universal Storage. The VAST Element Store manages all the SSDs in a VAST Cluster to form a single pool of storage and manages that capacity as a single namespace accessible as a file system and/or object store. The VAST Catalog is an extension of the Element Store, making it possible for VAST clusters to catalog every file and object.

The VAST Catalog will enable several new applications of VAST OS:

VAST administrators will leverage the Catalog for capacity management & chargeback
Backup and archive applications can use a new differential to traverse the namespace even faster, resulting in faster backups and the application of rapid data migration
Applications can replace POSIX functions with SQL statements to see rapid accelerations for POSIX operations.

Spotlight on Policy-Based Quality of Service

Customers get an alternative to the typical tiering mechanism and logical thin provisioning VAST hardware pools into different service classes. This release extends VAST Data QoS offering beyond the Pools concept by offering the ability to create additional service plans that sets min/max limits on bandwidth and IOPS per each VAST User or View (VAST’s term for a share, export, or bucket) such that public and private cloud service providers can contain noisy neighbors.

Customers that still need a tiered solution to offer application owners have the option to create different service classes that can be provisioned (and changed) from one simple-to-manage and scalable cluster.

Spotlight on Enterprise Key Management

Keeping with the service provider focus, VAST Data supports KMIP 1.2 for external key management. This Key Vault interface API allows users to connect EKM systems to VAST Clusters, adding the capability to manage unique encryption keys, the ability to rotate keys regularly, and control encryption methods while still saving on infrastructure. VAST Data also supports Thales CypherTrust and IBM Key Protect, with plans to expand the overall support matrix.

Spotlight on Global Snapshots and Global Clones

VAST’s Element Store maintains a granular approach to implementing Snapshots. Snapshots are reserved without affecting application performance with a byte-level granularity that allows customers to take 100,000s of snapshots at any namespace depth. Each VAST server accesses the same global dataset without worrying about partitions or volumes because of VAST’s Disaggregated, Shared-Everything (DASE) Architecture.

With the new software release, VAST clusters support the ability to share and extend snapshots to multiple remote clusters. Remote sites can most another site’s snapshots and make clones of the data to turn a snapshot into a read/write View. This new feature lays the foundation for other innovations relating to building a global namespace from edge to cloud.

Spotlight on Uplink with ML-Informed Capacity Prediction

Uplink is a global, cloud-based fleet management console designed to manage VAST clusters from a single pane of glass. With the power of elastic computing and machine learning libraries, customers are more informed about data utilization trends within the clusters and can take advantage of Capacity Predictions to stay ahead of demand.

Uplink now provides a one-click (tunable) prediction basis and a prediction horizon built on a time-series database made available as a capacity estimation tool. ML helps the system understand how to normalize against data trend anomalies that would otherwise confuse a classic statistical model, creating more accurate capacity predictions.

Spotlight on More Zero-Trust Security Work

In addition to supporting KMIP/EKM, VAST Data is extending the Zero-Trust agenda with support for TLS-based access for NFS, PseudoFS, Rocky Linux for VAST OS, and S3 Audit.

Spotlight on VAST’s Kubernetes CSI Update

VAST’s CSI now supports different storage classes for persistent and transient volumes. Helm Charts are provided for a more straightforward implementation of VAST’s CSI, adding simplicity to power on-prem and private computing clouds.

Engage with StorageReview