In a recent conversation with a friend in AWS I was reminded of the very interesting capabilities of AWS Snowball and the even more capable Snowball Edge. Snowball is a data transport solution that is designed to get large quantities of data from one place to another. It can move petabytes from legacy datacenters into the cloud, for example. Or from the cloud to off site systems if you would ever have a reason to do that. This original use case of moving data is still its most critical contribution, but the appliance can now be used to execute workloads over data. What this means is that data can be collected remotely, some analysis done over that data, and then moved into the cloud. This will be increasingly important in the coming age of industrial automation and enterprise digitization.
What Is The AWS Snowball?
Snowball was first a solution to data transport problems. If you are moving petabytes of info to the cloud, moving it via network connections just doesn’t make sense. For example, if your business has a T3 line and you want to move 100TB, it will take 269 days to get that job done. Meanwhile, many businesses are already dealing with 100’s of TB that they want to leverage in AWS, and some are dealing with petabytes.
To move the data, you order the Snowball device and it is shipped to you, you connect it to your systems, you manage the transfer of data to it, the device ensures the data is encrypted, and you ship the device to Amazon.
Each Snowball can hold 80TB. It weighs less than 47 pounds (more when full of data (just kidding, that was some CTO humor there)).
What Is The AWS Snowball Edge?
Snowball Edge is evolved from the Snowball and has expanded, more incredible features. It is still able to move data into and out of AWS (it holds more data than the Snowball, up to 100TB). But the Edge can undertake local processing/edge-computing workloads.
The device is designed to withstand and operate in harsh environments with no need for cooling and has also been shock tested. It has passed all the certifications you would expect of a well engineered enterprise technology, including being HIPPA certified. It has been rated by the government as able for use with data at the TS/SCI level.
Use cases include local storage and compute. You can use one or even a cluster of these devices to leverage AWS compute services (like AWS Lambda) in environments where there might not be Internet connections. This is giving new, on-premise data management/analysis capabilities. This will come in handy for organizations where large quantities of data are used and need to be analyzed closer to the point of collection before moving to the cloud. For example, in IoT or Industrial IoT situations where data might need to be rapidly looked at for analysis/conclusions and then shipped to the cloud for more in depth analysis/storage/assessment.
What comes next with Snowball Edge? If history is our guide, AWS is not finished innovating here. I’ve heard tidbits I can’t repeat, but for now would just say everyone should use their imagination and stand by to be blown away again.
By the way, the Snowball Edge is ruggedized. Here is evidence, a military standard, well instrumented barge test:
Here is a good use case that shows the Edge in action:
I remember the old days, just a few years ago. if someone wanted to do this type of thing they would invest in large network attached storage devices, configure them, ship them, load all sorts of tools on them, and $10 million dollars later have a solid enterprise system. What is the cost of Snowball and Snowball Edge? So low it is a rounding error in any IT budget. A typical job may cost $200.00.
Anyway, congrats to the engineering team at AWS. You guys are making magic.
Latest posts by Bob Gourley
- Announcing OODA: A Company Addressing Next Generation Security Challenges - January 9, 2019
- An Interview with Greg DeArment, Head of Infrastructure at Palantir - December 17, 2018
- Senzing: AI powered entity resolution to find who is who and what is what in your data - December 12, 2018