Hammerspace Architecture

Universal Data Access Layer
Metadata Control Plane
Data Orchestration
Automated Data Services
Expansive Storage Options

Universal Data Access Layer

Hammerspace enables any application and any user to share the same data, no matter where it is stored.

Multiprotocol Access:

Hammerspace supports multi-protocol shared acceso to the global data environment. Users and applications connect via NFS, pNFS, SMB/CIFS, and Kubernetes PersistentoVolumes. With cross-protocol data sharing, the same data can be accessed from Windows, Linux, macOS, UNIX, VMWare, and Kubernetes environments.

Users and Application Anywhere:

All data may be viewed and has high-performance read/write access through any protocol, regardless of the protocols supported by the underlying storage resources.

High-performance:

Hammerspace delivers high performance parallel file system performance without the need for proprietary clients. Hammerspace leverages the native pNFS client built in to Linux to deliver the high-performance needed for HPC, M&E post production, and other demanding use cases.

Cross-protocol data sharing enables read/write of all data through any protocol

Hammerspace provides a CSI (container storage interface) driver supporting shared-block, file-backed (local file), and shared file (NFS) persistent volumes individually and in mixed mode.

Hammerspace enables any application and any user to share the same data, no matter where it is stored.

Multiprotocol Access

Hammerspace supports multi-protocol shared access to the global data environment. Users and applications connect via NFS, pNFS, SMB/CIFS, and Kubernetes Persistent Volumes. With cross-protocol data sharing, the same data can be accessed from Windows, Linux, macOS, UNIX, VMWare, and Kubernetes environments.

Users and Application Anywhere:

All data may be viewed and has high-performance read/write access through any protocol, regardless of the protocols supported by the underlying storage resources.

High-performance:

Hammerspace delivers high-performance parallel file system performance without the need for proprietary clients. Hammerspace leverages the native pNFS client built in to Linux to deliver the high-performance needed for HPC, M&E post production, and other demanding use cases.

Users Have Local Access to All Data

User View

Meta Data File System

Global metadata control provides global data access to users and applications, with an experience the same as if they were accessing a local primary NAS share. Hammerspace enables custom metadata tagging to makes it easy to better describe, classify, and manage the orchestration of file data based on your organization’s business needs.

Metadata Assimilation:

What was previously multiple otherwise incompatible storage types with individual file shares, can now all be aggregated into the Hammerspace global file system. Users still only see the files they have permissions to see. File system access is now global, bridging any local, remote or Cloud storage resource from a common file system view.

No Data Migration Required:

Adding existing data into Hammerspace is made easy with data-in-place metadata assimilation. Data does not need to move anywhere. Hammerspace simply scans the share on your existing storage platforms and aggregates all metadata about the files and storage resources, almost instantaneously,

Custom Metadata

Metadata tagging ensures files are appropriately classified with critical information without relying on user action. This powerful ability provides enormous value and automation for a number of use cases, including locality of data, data migrations, data protection, disaster recovery, active archiving, rendering jobs, simulations, data analytics, and more. Automatically tagging datasets with department ID, job reference or other business data makes cost allocation for charge back or show-back a simple task, even though the files for those Departments may be stored across multiple storage platforms and locations.

Hammerspace can take over a network share of any size and begin serving read/write requests immediately.

Customize your metadata to organize and utilize data across your enterprise

Global metadata control provides global data access to users and applications, with an experience the same as if they were accessing a local primary NAS share. Hammerspace enables custom metadata tagging to make it easy to better describe, classify, and manage the orchestration of file data based on your organization’s business needs.

Metadata Assimilation:

What was previously multiple otherwise incompatible storage types with individual file shares, can now all be aggregated into the Hammerspace global file system. Users still only see the files they have permissions to see. File system access is now global, bridging any local, remote or Cloud storage resource from a common file system view.

No Data Migration Required:

Adding existing data into Hammerspace is made easy with data-in-place metadata assimilation. Data does not need to move anywhere. Hammerspace simply scans the share on your existing storage platforms and aggregates all metadata about the files and storage resources, almost instantaneously,

Custom Metadata

Metadata tagging ensures files are appropriately classified with critical information without relying on user action. This powerful ability provides enormous value and automation for a number of use cases, including locality of data, data migrations, data protection, disaster recovery, active archiving, rendering jobs, simulations, data analytics, and more. Automatically tagging datasets with department ID, job reference or other business data makes cost allocation for charge back or show-back a simple task, even though the files for those Departments may be stored across multiple storage platforms and locations.

Hammerspace can take over a network share of any size and begin serving read/write requests immediately.

Customize your metadata to organize and utilize data across your enterprise

Metadata Assimilation:

 

Object Based Policies:

 

Data Orchestration

Optimize storage resources with a machine learning policy-engine. Data orchestration is policy driven and users have the option to self-service movement of data to the location, application, or compute environment where it is needed. The global file system automates and optimizes the management of data across clouds and sites. With live data mobility, non-disruptively move data, even while it is being read/written to, between storage systems, sites, and clouds.

Global Namespace

Hammerspace creates a single logical system that spans storage infrastructure and sites regardless of distance. This is not System A moving data to System B and System C in the way that legacy storage systems operate. Instead, this is System A, B, and C’s metadata existing in all of your locations, which results in simplicity of viewing data and moving to the appropriate location when needed.

File-Granular Replication

Data is managed at file-level granularity efficiently delivering scale across complex mixed, distributed infrastructure without creating unnecessary copies of entire volumes of data.

Objective-Based Policies

The extensible metadata describes is referenced by user objectives which is are coupled with machine learning driven automation to move and tier data across storage, sites, and clouds. Hammerspace provides intelligent policies to orchestrate and manage your data. You can create policies ascribed to any metadata attribute including file name, file type, owner, path, create time, modify time, etc. Metadata is customizable and programable, allowing you to add your own labels within the metadata and then apply policies to based on the enhanced metadata.

These objectives are implemented either through the GUI, or via CLI, or REST API. The GUI provides a variety of checkboxes that instantly determine attributes such as availability, durability, placement, protection, and a host of other options. The same objectives can also be realized through the Hammerspace CLI, or Hammerspace Toolkit, which is downloadable through Git Hub. For example, you could group all files associated with a particular project under a specific name and add metadata called “Project ABC” with specific policies for anything related to that project.

Automated Data Services

Hammerspace provides file-granular data services leveraging its policy engine or on-demand capabilities. File-granular services give individual files, or sets of files, the ability to be managed by policies for any metadata attribute including file names, creation dates, modify times, file types, etc. Hammerspace data services enable companies to manage their data-driven business elements in ways that were previously impractical, or even impossible, due to price and performance challenges.​

Data Services

Meet compliance, data protection, disaster recovery, and ransomware protection requirements.

Expansive Storage Options

Unify data residing in legacy DAS, NAS, SAN, object storage, and cloud storage into a global data environment.

Leverage a single global file system across any storage infrastructure, in any cloud simply by replicating only the metadata across all sites. By managing data through the metadata users gain universal visibility and insights from enterprise file data across the entire infrastructure while eliminating data silos. This further allows for capacity, performance, and scale optimizations only possible with granular controls and total automation. Eliminate disruptive data migrations with live mobility

Utilize Existing Storage Infrastructure​

Utilize Existing Storage Infrastructure​
Hammerspace can utilize the available capacity of your existing storage infrastructure, reducing cost and time-to-value. ​ You can get up and running with Hammerspace without having to acquire any new storage infrastructure.​ When new storage needs to be added, Hammerspace can deploy new storage bare metal, virtualized, or in the cloud.
Universal Data Access Layer
Metadata Control Plane
Data Orchestration
Automated Data Services
Storage
Universal Data Access Layer
Metadata Control Plane
Data Orchestration
Automated Data Services
Storage