Hammerspace Introduces an All-New Storage Architecture | Read the Press Release

How To Create a Global Namespace With Hammerspace

Hammerspace is a software-defined data orchestration and storage solution that provides a single, global namespace across all storage in data centers and different clouds.  

Once granted appropriate permissions and access, users and applications connect to the Hammerspace namespace and can view, access, and modify data in any storage system and in any location without creating copies of data or involving IT to connect to each storage system.

Designed to make data a global resource across distributed or otherwise incompatible storage platforms, Hammerspace presents a cross-platform global namespace where users and  applications can have direct, multi-protocol access to all files, regardless of which storage type, cloud vendor, or location they are in today, or move to in the future. 

The metadata layer is common across all users, everywhere. In other words, all users access the same file metadata regardless of where the files are actually stored, without the need to manage file copies between silos or locations. If a local instance of a file is needed for processing in one part  of the world, orchestration of a file instance is a background operation that is transparent to the users or applications. The metadata layer of the file system is still the same across all silos and locations. This means users everywhere are accessing the same file metadata, not forked file copies.  

Why does a global namespace matter? 

Modern compute infrastructure, whether on-premises or in a public cloud, is often limited by difficulty accessing distributed data sets. Hammerspace enables efficient global use of compute resources regardless of where they are, or where the data resides, allowing customers to focus on actually using their data rather than wrangling it across data silos.

A key benefit of Hammerspace’s technology is that it protects file access by users or applications from changes to storage infrastructure, or from data movement between different storage types and locations. This enables always-on global file access to users and applications via standard protocols with persistent mount points, regardless of which storage types the data moves to, or what location users and applications are accessing the data from. 

So, how do we do it?  

  • Data-in-Place Metadata Assimilation 

This refers to the rapid harvest of file system metadata from existing storage systems, which is done without the need to migrate the data to another storage location. 

Even very large environments can be rapidly assimilated, so users and applications can begin browsing content through the global namespace within minutes, even as additional file metadata is being harvested in the background. 

This also enables customers to extend the life of existing storage investments, effectively extending them to include additional storage types and the cloud, rather than fragmenting access with new silos of additional storage, or needing to replace them prematurely.

  • Objective-Based Policies

Objective-based policies may be established within Hammerspace for orchestrating data at a file-granular level – completely in the background – for tiering, platform migration, data protection, workflow provisioning and much more. Users and applications retain the same view to the global file system regardless of changes at the infrastructure layer or data movement between systems or locations over time. 

  • Global Data Services

Critical data services may be implemented globally across all storage types and  locations, for data protection and other use cases, without the complexity or fragmentation often associated with limited point solutions, gateways, or other vendor-locked techniques. 

Getting started

Once you have decided to move to Hammerspace to unify your  data across storage silos into a single namespace, here’s how the process will work: 

Assessment

  • Determine the data storage silos you have and where they are located.
  • Identify the data sets that you want to unify and their respective access patterns.

Deployment

  • Deploy the Hammerspace software according to your platform (on-premises, cloud, hybrid). Depending on the architecture, you might need to deploy Hammerspace nodes on each location where your storage resides.

Data Ingestion

  • Assimilate (onboard) your storage silos into Hammerspace. This step involves adding your storage resources (such as NAS, SAN, or cloud storage) to Hammerspace so that it can start managing the data on those resources.
  • Create the global namespace. This namespace will act as the single point of access for all your data.

Data Management

  • Use Hammerspace’s metadata-driven orchestration to set policies for your data. These policies can define how and where data is stored, replicated, tiered, or protected.
  • You can also leverage its machine learning capabilities to automate some of these operations based on the actual usage patterns of your data.

Access Control

  • Configure user and group access rights to ensure that only authorized personnel can access the data.
  • Hammerspace can integrate with enterprise identity providers to simplify access control.

Monitoring and Reporting

  • Utilize Hammerspace’s analytics and reporting tools to keep track of your data’s performance, usage patterns, and more. This can help in optimizing costs and performance.

We recognize that each organization’s data landscape and needs are unique. We will work with you through each step of the process to ensure that your setup and usage of the platform align with your organization’s goals and requirements.

Learn more by downloading the Hammerspace Technology White Paper.

About the Author

Floyd is Vice President of Product Marketing for Hammerspacer. He has been involved with data management and storage for more than 25 years, focused on the methods and technologies needed to manage extreme volumes of data to keep up with the needs of modern, distributed storage resources and workflows.