What Is Edge Storage?

What Is Edge Storage?

Edge storage is data storage placed closer to users and data sources at the edge of a network to reduce latency and network data loads.

How Did Edge Storage Come About?

Edge storage architecture has evolved over time—first to serve widely dispersed workforces using edge drives for individual users, then for local network-attached storage devices for wirelessly connected devices. Today, edge storage is an essential part of computing being performed at a whole new edge: that of IoT devices, remote data collection, autonomous vehicles, and mini data centers geographically dispersed to deliver better performance and enable new use cases.

Why Is Edge Storage Important Today?

A technology convergence that’s been years in the making is currently underway, and edge storage is playing a key role. Progress in high-speed connectivity (especially 5G), cloud computing, AI and machine learning (ML), and other technologies is enabling new possibilities for edge computing—from autonomous devices to new data streams to new mobile applications. In each case, high-performance storage has a role to play. For example, edge storage might contain: 

• Customer information to ensure a seamless low-latency experience
• Data collected by a network of sensors or cameras
• Imagery used by a self-driving vehicle
• Resources needed to enable high-performance mobile gaming

What Technologies Does Edge Storage Use?

Data storage technology has played an important role in the evolution of edge computing. Today’s deployments of edge computing typically use solid-state drives (SSDs), which are reliable, high performing, and easily scalable to the most demanding use cases. 

PCI-based SSDs that use non-volatile memory express (NVMe) perform even better and are particularly suited to data-intensive tasks such as remote data collection involving images. Even removable SSD media have high enough capacity—up to 128TB currently—for use cases such as video surveillance.

Benefits of Edge Storage vs. Cloud Storage

Since the benefits of edge storage overlap with those of edge computing—including low latency, network optimisation, lower costs, and new use cases—it’s useful to compare edge storage to cloud storage, another choice offered when implementing edge computing. Every organisation’s needs are different, but generally, edge storage has the following advantages over cloud storage:

• Lower cost: Physical storage deployed on the edge requires a greater initial cost compared to cloud storage, but over time, the cost of physical storage will be lower than cloud storage and bandwidth that’s paid for on a monthly basis. In addition to reducing cloud footprint, edge storage lowers costs by enabling workflows and automations that reduce bandwidth requirements.
• Lower latency and higher performance: For many use cases, cloud architecture simply introduces too much latency to support the performance demanded by end users and advanced applications. The very large data loads of some use cases, like image-based data collection, wouldn’t just take a long time to move from place to place, they might also overload networks needed by other users. By caching data and processing at the edge, sampling the data, or sending abbreviated versions, edge storage saves time and prevents the very large data loads of some use cases from overwhelming other networks within an organisation.
• More control over data: Some organisations may need to maintain physical control over data at all times, such as for regulatory compliance or to eliminate security risks. And while some may find the cloud more resilient and agile, others may prefer to have direct control over every aspect of their edge deployments, including storage, and to steer clear of dependence on a cloud vendor in the event of an incident.

Edge Storage Use Cases

Potential use cases for edge storage are vast. Let’s look at a few sectors where it’s playing an important role. 

Agriculture

Edge storage is used to collect, aggregate, and store the enormous amounts of data generated by environmental sensors and cameras spread over large agricultural areas. By enabling data enrichment at the edge, edge storage helps growers get insights more quickly.

Manufacturing

Factories use AI to perform precision monitoring of processes and systems to ensure quality and consistency. Edge storage is used to store the data that AI uses to train its models.

Transportation

Autonomous vehicles need to aggregate and process data from multiple sources to operate safely and effectively. Sending data to a remote data centre isn’t an option for rapid responses, so autonomous vehicles use edge storage and edge computing to perform calculations in real time.

Security

Security cameras create extremely large amounts of data, nearly all of it irrelevant for later use. By storing this data at the edge and training an edge computing system to only send anomalous footage or timed samples, analysts gain needed information while making more efficient use of network resources. Purpose-designed edge-grade SSDs enable the exhaustive amounts of writing and rewriting of data involved in such systems.

Pure Solves Your Edge Computing Challenges

Thinking about bringing in your own edge storage solution? Keep the roadblocks in mind before you move forward. Edge solutions are usually created for a specific need, which means they may not have the resources and capabilities for different projects.

To make sure your edge solution meets your organisation’s particular needs, spell out the scope of your project in as much detail as possible, including:

• The minimum level of connectivity required
• Data life cycles needed
  
• Plans for security

Pure Storage® has several solutions for meeting your edge storage needs:

• FlashBlade®: As the leading unified fast file and object (UFFO) platform, FlashBlade is suited for analytics, machine learning, artificial intelligence, and similar edge data processes.
• FlashArray™: The all-flash storage pioneer and leader, FlashArray combines the high performance of all-flash storage with VMware integration to create a hybrid-cloud solution. FlashArray supports a 5G multimode infrastructure combining virtual machines and containers.
  

The high-density, low power consumption and easy remote management of both FlashBlade and FlashArray make them well-suited for challenges arising at edge locations.

• Portworx®: The most complete Kubernetes data services platform, Portworx provides a storage layer for running cloud-native workloads at the edge. Portworx delivers a complete solution for containerized workloads, including backup and disaster recovery. Portworx integrates with FlashBlade and FlashArray for high performance and reliability.