The early buzz around AWS Greengrass revolved around how it brought AWS functionality to the internet of things....
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
But Greengrass offers more than just IoT capabilities and, for some organizations, has led to a new way of thinking about hybrid cloud architectures.
Greengrass extends the AWS security model to a wider range of enterprise applications and opens the gates to more distributed applications across hybrid cloud architectures. Developers can use Greengrass Core, which provides local services to devices, to expand Lambda functions out to connected IoT devices. Enterprises can also use Greengrass to aggregate, filter and summarize large data sets from IoT devices inside an AWS cloud ecosystem. To capitalize on these capabilities, AWS integrated Greengrass Core software into Snowball Edge hardware to create a dedicated IoT gateway.
Snowball Edge provides a limited set of compute functionality to run AWS Lambda functions offline. The appliance also enables enterprises to cost-effectively perform large data transfers into the cloud -- as many as hundreds of terabytes of data. In addition, enterprises can effectively rent Snowball Edge to protect sensitive data in hospitals, factories and other settings.
Greengrass' impact could go beyond simple IoT device management toward a new style of hybrid cloud, where more data is processed in the field. And, together, Greengrass and Snowball Edge help push AWS technology, such as Lambda functions, out to local resources to expand the boundaries of hybrid cloud architectures.
Greengrass goes beyond IoT devices
The Greengrass ecosystem consists of Greengrass Core software, which users can install on edge computing gateways or servers, and the AWS IoT software development kit for optional connected devices. By essentially placing AWS functionality and security capabilities onto private infrastructure or devices, Greengrass enables enterprises to work with and expand their hybrid cloud architectures.
Two key aspects of Greengrass security relate to data protection and app management. Greengrass relies on AWS Identity and Access Management to update Lambda functions from the cloud. This security provision requires certain access privileges to run update code, and it prevents hackers from directly accessing the device.
Snowball Edge further encrypts all data stored on the hardware. Without Snowball Edge, enterprise IT teams need to add at-rest data encryption to secure Greengrass Core servers.
Multitask with Snowball Edge
Snowball Edge enables enterprise architects to provide high-speed storage and limited data processing -- comparable to an m4.4xlarge instance -- closer to devices. Ops can work with it offline or manage it remotely, as long as it's occasionally connected to the network, and they can also deploy new security certificates over low-bandwidth mobile networks to Snowball Edge devices when required.
Snowball Edge early adopters
AWS Snowball Edge provides onboard compute for Amazon's high-capacity storage appliance. The device can help teams do work in the field, even with limited cloud connectivity. Here's how some organizations use Snowball Edge:
- Oregon State University deploys Snowball Edge devices on oceanographic expeditions to replace the dozens of hard drives it formerly used to record scientific data from sensors. Lambda functions running on these devices help reduce the amount of data. The team mails devices back to AWS for deeper analysis in the cloud.
- The Electric Daisy Carnival deploys Snowball Edge devices to its week-long festivals with 130,000 attendees to help aggregate the collection of terabytes of videos and pictures captured by employees and attendees. Snowball Edge enables a staff to access, tag and prune this media securely at the festival, and they later send it to AWS for longer-term storage and archiving.
- Philips integrated Snowball Edge into its HealthSuite digital platform, which enables smaller hospitals to securely manage large data sets collected from data-intensive equipment, like MRI machines. Hospitals can store this data and then ship it to the AWS cloud for longer-term archival.
Snowball Edge pricing starts at $300 for 10 days and then $30 per day after that, which comes out to $10,650 per year. While that number seems high, it could be worth the price to avoid the impact of possible data breaches. Early uses of Snowball Edge include offline video and image collection, scientific data aggregation and healthcare.
Greengrass expands hybrid possibilities
AWS says that Greengrass can enable a higher set of use cases than Snowball Edge alone with more programming flexibility and integration with different kinds of hardware. In many cases, Greengrass will also be cheaper than Snowball Edge, unless you plan to send terabytes of data to the cloud.
Greengrass Core essentially provides the ability for enterprises and third parties to package some of the best aspects of AWS security and management into a Greengrass service that costs less than a quarter per month, plus an annual fee. Third-party hardware vendors, like Dell and IBM, are starting to incorporate software from providers like Canonical, a proponent of open source software, to stake their claims to the edge computing market. And once users establish best practices for these deployments, consultants will promote similar hybrid approaches.
Enterprises and third parties can customize various kinds of edge servers with more Lambda functions. Over a dozen companies, including Qualcomm Technologies, Intel, Raspberry Pi Foundation, Samsung and Lenovo, plan to embed AWS Greengrass into third-party hardware.
Greengrass deployments reduce latency between apps running on edge servers and IoT devices or existing enterprise data stores. Nokia reported that early experiments found round-trip times decreased 28% and latency reduced 39% by processing data at the edge compared to the cloud. This kind of distributed architecture could reduce the time to transform, summarize or analyze enterprise data stored on private servers before it's sent to the cloud.