How to make data globally usable regardless of where users and data are.
Hammerspace enables enterprises to use all of their data with any application, any user, any data center, or cloud service, anywhere.
The 41st IT Press Tour had the opportunity to meet with David Flynn, CEO, Floyd Christofferson, VP of Product Marketing, Douglas Fallstrom, SVP of Products & Operations, and Johan Ballin, Director Technical Product Management at Hammerspace to learn about their Global Data Environment and the benefits of a Global Data Platform.
The Problem
Enabling local data access and data services across multiple local, cloud, and remote storage resources.
No one storage type serves all requirements. Data outlives the storage it’s on today. Different cost and performance bands are difficult to decipher. Different protocols are required for broader access. Incompatible silos emerge. There is increased IT complexity to manage. Even today, most storage systems are architected for local access.
Distributed storage and users add problems. Distributed workforce needs direct online access to data. Global management of data services, data protection, and intelligent data placement is nearly impossible. Cost and IT complexity grow. Data silos are no longer only local problems.
The Need
Data and storage management needs to adapt to today’s reality. Your customers are global, your team is global, your data needs to be global. Users require high-performance global access to data from any data center, any cloud, any region, anywhere. Global management and control of data and storage resources are a must.
A global data environment for hybrid cloud and multi-data center NAS requires a Global Data Environment to unify all data in a single global file system and provide universal data access and orchestration.
Global Data Environment
A Global Data Environment (GDE) provides users and applications with local read/write access to all of an organization's file data, which they have been granted a privilege to access, no matter where the data is stored or where the user or application is located.
The essence of creating the GDE is in creating metadata-driven workflows that provide a unified view of, and local access to, data for users, applications, and compute environments located anywhere in the world. The primary interaction point with data is first with metadata, and secondarily, data is moved only when required by policy or on-demand. When data is moved, it is replicated on a file-granular basis to ensure the efficiency of network usage and speed of access to the required data.
This allows admins and users to see data that is universally accessible. Hammerspace achieves this by taking the metadata and making it into a control plane to sit in front of the file system. The data moves behind the file system. Metadata is not replicated.
If your data management is above the file system, it is interrupting the perception of the data. It needs to be behind the file system to enable seamless operation. It's important to separate the metadata from the infrastructure.
Global Data Platform Components
Universal Data Layer Access. Supports NFS, pNFS, and SMB, individually and in mixed protocol mode. All data may be viewed by any protocol regardless of the protocols supported by the underlying storage resources.
Flexible Data Orchestration. Move data live and non-disruptively within the data center, across multiple data centers, to the public cloud, between different regions within the public cloud, and across heterogeneous clouds. Live data orchestration of individual files or entire shares to any location enables automated workflows using objective-based policies and is transparent to users or applications.
Automated Data Services. File-granular global data services across all local and remote storage resources leveraging the 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, in addition to custom metadata tags.
Expansive Storage Options. Use the available capacity of existing storage infrastructure, reducing cost and time-to-value, and even extending the life of current storage. This provides the ability to get up and running without acquiring any new storage infrastructure. When more resources are needed, it can be seamlessly added to the GDE.
Use Cases
Distributed Workforce – Single data set is accessible to all users and applications – fast, efficient, and fully automated.
Cloud Burst for Compute – Leverage limitless core count, on-demand with no data migration. Non-disruptive, continuously optimized, live data mobility.
Active Archive and Cloud Tiering - Tier data based on activity or organization requirements to the data center or cloud object storage easily and quickly.
Data Management - All users and applications have read and write access to the same network shares and data regardless of the location. File-granular replication ensures that only the data that is needed in each location is replicated to the appropriate sites.
Data Protection - Built-in data protection tools ensure data is protected and compliance requirements are met. As data ages, policy-driven tiering moves data to lower-cost storage to reduce cost, free up expensive infrastructure resources, and ensure proper data protection.
Disaster Recovery - If a site or system goes down, no failover or fallback procedure is needed since metadata exists in multiple locations providing continuous online access.
Multi-Cloud - New data services for compute resources, application processing, machine learning, and data storage can be created on a continual basis. Live and transparent data mobility across cloud providers, and cloud regions, make it easy to leverage the cloud service or application that is most suited for the job.
Ransomware Protection - Undelete feature automatically saves a copy of deleted files making it easy for companies to recover the most recent version of their files as they were prior to an attack. For ransomware attacks that do not delete data, file versioning stores the latest versions of the files so none are lost. A ransomware attack that may start to encrypt active data is mitigated by the ability to turn back the clock to a previous unaltered version of the file prior to the attack.
Storage Standardization - The platform is software-defined and can be deployed on a wide range of infrastructure including block storage, NAS, object stores, and cloud from different vendors. Users pick the deployment method that best suits their needs.
Case Studies
A global video effects production company needed to accommodate a distributed workforce of artists across five UK studios and global locations need “local access to their data. There was also a need to reduce cloud render costs. Remove render workload between cloud regions on-demand to get better rates. With the Hammerspace solution, artists can log in as local from anywhere via HS Global Data Environment. Artists have access to their work without making copies. Workloads can follow the sun and dynamically shift without disruption. Render workloads can move to the most cost-effective region seamlessly. Customer apps distribute render jobs between the US East and West, India, and Europe dynamically. Artists apply custom data tags to files at their desktops, simplifying downstream orchestration and processing. This results in reduced costs, eliminates Rsync copies and complex manual job orchestration, and increases the productivity of the artists.
Global distributed production environment, a gaming development and production company needed to manage data across eight data centers and two cloud providers. Data placement has been manual vi gateways and manual copies. They were having trouble supporting a globally distributed workforce. While talent is everywhere, data access is not. Hammerspace provided global data orchestration across all sites and clouds. Copy management is controlled and data placement is automated. Unified local access with the global data environment give artists anywhere the ability to access data workloads as local. This solution eliminated manual copies, gateways, and point solutions for data orchestration. Local access to remote data accelerates production and reduces delays.
Key Takeaways
Solves the problem of providing local access to distributed data.
Aggregated metadata powers the Global File System.
Enables unified and performant local file access to global storage resources across incompatible storage types and locations.
Reduces IT complexity with global policy-based control of data services across any storage type and location.
Customers reduce storage costs with:
Better utilization of storage resources, extending the life of existing storage.
Reduced OPEX for data and storage resource management.
Increased user productivity with ubiquitous local access to all files and workflows regardless of where they reside.