If you’ve followed along this December, you’ve seen the trials of the North Pole’s infrastructure migration. But now that the holidays are over, let’s take a more positive look at the changes that Santa and the elves were able to make. In the final part of our series, we’ll examine the North Pole’s new infrastructure, and how it helped them deliver more presents than ever this year.
The New Workshop
Santa’s infrastructure now starts with a brand new Christmas Wishes Pipeline. When letters to Santa come in (still via normal post, but rumor has it they’re working on a web interface for 2018), AWS Rekognition uses text extraction to parse the often sloppy handwriting into machine-readable data. The data is then sorted and placed into a massive S3 data lake that contains more than one billion records. These include each child’s name, location, and the items on their wishlist.
The Naughty and Nice List has also seen some serious upgrades. It’s finally gone digital and is now contained in a managed RDS database. Admins no longer need to run out to the stables to manually check each child’s status – they can run a simple query and make updates as needed.
Speaking of updates… regularly scheduled Lambda functions now run every five minutes to query the S3 data lake using Athena. Depending on the return value, checked against the Naughty and Nice List, they may update each child’s delivery status. As a result of this automation, manual admin queries are down by nearly 80%. Thanks to AWS infrastructure, the Naughty and Nice List practically runs itself.
On Christmas Eve, the final delivery list is stored in a separate S3 bucket – the Deliveries list. Each good child’s name goes into this dataset, and from there, Santa’s delivery route is calculated to save up to 10 tons of sleigh fuel(reindeer food) per year.
This leads us to the next stage of the system: big data. It’s no secret that most companies are using advanced analytics for business, and Santa’s elves wanted in on the action. A machine learning algorithm now calculates his delivery route based on a number of proprietary factors. Over the next decade, this is expected to cut delivery times almost in half.
The manufacturing line has also seen major changes. Toy production machines still run on elf magic, but this is being supplemented with thousands of IoT sensors that can immediately detect a component failure. This data is processed and sent to QuickSight, which means the days are long gone when Santa would have to make hourly rounds on the workshop floor. Instead, he can now monitor production from his office while sipping hot cocoa.
The gift wrapping machines were similarly outfitted with IoT sensors. However, because they’re more prone to error, their backend is a bit different. CloudWatch is used to monitor the incoming Kinesis stream of data for paper jams or failures in the bow-tying mechanism. These mechanisms still require manual involvement when they break, so an SNS topic is set to notify an on-call engineer when intervention is needed. With these changes, the workshop only requires a single on-call elf at any given time – the rest are free to focus on stuffing stockings and feeding the reindeer.
The reindeer haven’t been left out of the excitement either. Cloud migration is a huge undertaking, and it doesn’t happen all at once. The elves keep several racks of servers in the stables as on-prem backups, and as a result, have finally added climate control (Rudolph was glowing, literally, as he told us about this development). Santa has also invested in several GameLift servers for use during downtime, and after all these years, multiplayer reindeer games are finally a reality.
Security has been given major attention. Because all the North Pole’s data and most of its workloads now live in the cloud, each service requires multifactor authentication and access is closely guarded by IAM roles. Are the threats entirely eliminated? Of course not. But when a DoS letter comes in from the Easter Bunny, or when the children of billionaires hire hackers to switch their names to the Nice List, Santa can now rest easy knowing that some basic measures have been implemented that will cover the vast majority of attack vectors.
See Santa’s new workshop infrastructure in its entirety:
This year, deliveries were 10% faster. Santa delivered 30% more presents, the margin of error on incorrect deliveries was practically negligible, and best of all, the elves once again have confidence in their operation. With results like these, it’s hard not to get excited, and it all started with the right training.
Cloud platforms like AWS are excellent tools for business, but that’s exactly what they are – tools. To get results, employees (and elves) need to know how to use them, and that’s exactly what we were able to help Alby Snowball and his team accomplish this year. People can do amazing things when they have the right skills. That’s why we focus on hands-on training, using real world scenarios to help people learn the technology they need to propel themselves and their businesses into the future. The North Pole’s cloud migration was a great story of student success, but it’s not the only one. Thousands of students have used Linux Academy and Cloud Assessments to upgrade their skills, enhance their careers, and reach their potential. As 2017 draws to a close, we’re happy to celebrate Santa and the elves’ success. But we also want to look forward to the future, and as 2018 approaches, we hope to share in celebrating yours.