Edge computing and cloud computing can work hand-in-hand, but there are times when their paths diverge. When it comes to storage, for example, it can be impractical to save the large amounts of data created at the edge directly to the cloud. Organizations’ edge computing storage strategy, therefore, needs to factor in a number of considerations.
How many different edge computing storage solutions will you need?
Just as an organization doesn’t have just one type of data, it likely won’t have just one edge computing storage solution. Edge computing storage solutions are commonly chosen based on how the data will be used, and sometimes this may involve using multiple storage devices.
Imagine that an organization has a collection of IoT devices that are collectively producing a lot of data. The organization ultimately wants to store that data in the cloud, but sending the raw data to the cloud directly from edge devices may be problematic.
The data that is produced by these devices likely far exceeds what the organization’s internet bandwidth can handle. Sending all of that data directly to the cloud may not be an option. Even if sufficient bandwidth does exist, it may be cost-prohibitive to store all of that data in the cloud. In these types of situations, edge storage is often used for the raw data. Data can be processed at the edge, reducing its footprint before it is sent to the cloud.
In a situation like this, the organization might configure the IoT devices to send their raw data to an on-premises device that can store the data while it is being processed. Once processed, the data could be sent directly to the cloud.
Some IoT providers are even starting to include on-device storage and processing capabilities.
When IoT devices were first introduced, they were intended to be lightweight and inexpensive. As such, the devices often contained the bare minimum hardware required for the device to function properly. It was therefore almost unheard of for these devices to include internal storage. Now, some manufacturers have begun introducing devices that include enough storage and CPU resources to enable the devices to process data internally. Once data has been processed on the device it can be sent directly to the cloud, or it can be sent to another device at the edge for additional processing.
What are your edge computing storage use cases?
Edge computing storage isn’t just about accommodating IoT devices. Sometimes, edge storage is used as a buffer that temporarily stores data before it is sent to the cloud or to an organization’s primary data center.
Say an organization is creating data in a remote location with slow or intermittent connectivity. In this type of situation, sending the data to cloud services in real time probably isn’t a good option. The organization may suffer data loss if the internet connection drops or slows to a crawl. A cloud storage gateway solution can temporarily store data at the edge until it can be sent to the cloud. A cloud storage gateway can be a physical machine or a virtual appliance, so long as it is equipped with storage.
At the most basic level, a cloud storage gateway temporarily stores data until sufficient bandwidth becomes available to send the data to the cloud. However, some cloud storage gateways do far more. It is relatively common, for example, for a cloud storage gateway to deduplicate or compress data in an effort to reduce the required bandwidth. Some cloud gateways also maintain an on-premises copy of any hot data, so users do not have to access that data directly from the cloud.
What are your edge computing storage requirements?
As an organization evaluates its edge computing storage needs, there are two main things that it needs to consider. The first consideration is the role that the storage will play with regard to the organization’s overall architecture. Will the edge storage be used solely as a data buffer? Will it act as a temporary repository for data that needs to be processed? Or does the storage simply need to make data locally available to users who are working at the location where the data is being generated?
For more on edge computing, read Omdia's New Compute Ecosystem: From Cloud to Edge 2021 Report, which offers an overview of the market for IT and physical infrastructure equipment deployed at edge locations as well as insights from 18 edge thought leaders.
The second important consideration is storage capacity. Capacity planning involves anticipating the amount of data that the edge storage device must accommodate based on the rate at which data is being created and the duration for which the data will need to be retained on that device. It is also important to consider any anticipated data growth. If, for instance, edge storage will be used for processing data generated by IoT devices, you will need to take into account any new devices you will be adding in the foreseeable future.
How will edge computing storage fit in with your overall storage architecture?
It is entirely possible that an organization may decide to store all of the data generated at the edge on edge storage. More often, however, edge storage is meant to be temporary, and the data (or some variation of the data if the data is being processed at the edge) ultimately needs to be sent to the cloud. As such, an organization will need a solution for securely and efficiently migrating the data.
The best option for migrating the data is going to depend heavily on the type and volume of data at the edge, and how that data is being stored. If for example, if the data is being generated by IoT devices, then the best option might be to use an IoT gateway to collect the data from the various devices, process the data and send it to the cloud.
End user data such as document files can be another form of data at the edge. This is particularly true if users are continuing to work from home and are saving data to their laptop hard drive. In this case, selecting an edge storage solution isn’t really an issue since the laptop’s hard disk is acting as edge storage. Instead, the organization needs to consider how best to move the data from a user’s laptop to one of its data centers or to the cloud. The best option for moving the data may be to adopt an enterprise file sync and share solution.
These are just a few of the considerations that an organization will need to think about as it develops an edge computing storage strategy. Other considerations include the cost of storing data at the edge as compared with storing it elsewhere, the security and compliance implications of storing the data at the edge, and the organization’s available bandwidth. If you find bandwidth to be an issue, it may even be worth considering if bandwidth prioritization through QoS or various data reduction technologies like deduplication or compression could help.