As the cloud keeps building momentum, it’s no surprise that the subject got a lot of air time at FutureStack: New York this fall. The four presentations highlighted here—from AWS, Google, New Relic, and Braze—cover everything from the future of multi-cloud to the role of New Relic Infrastructure in cloud migration to the power of auto-scaling.
Let’s take a look…
Reliable cloud migration
Despite its obvious benefits, adopting the cloud can be daunting, as Kevin Cochran, solutions architect at Amazon Web Services, acknowledged in his talk on Reliable Cloud Migration. Resisting change is part of human nature, he said, “but migration is as old as life itself.”
To make the case for cloud, Cochran laid out some of the drawbacks of managing on-premise servers: “You’ve got real estate, you’ve got utilities, you’ve got physical security that you have to deal with, you have staffing requirements.” Then there’s burn-in, maintenance, and the pressure to maintain redundant hardware in case of emergencies. All of which adds up to massive property-management headaches.
AWS, Kevin said, relieves those headaches. He outlined the benefits, from first-rate server and application security to near-infinite storage. With AWS Lambda, the serverless compute platform, you can throw together a Restful API in just 10 minutes. Elastic Load Balancing takes care of your fault tolerance needs, and AWS’s broad database offerings cover even more bases.
Watch Kevin’s full presentation here:
New multi-cloud integrations for New Relic Infrastructure
Regardless of which service (or services) you adopt, monitoring performance before, during, and after your cloud migration is essential to a successful adoption. According to Alberto Gomez, director of product management at New Relic, it all comes down to dynamism, change, and cost. “Today, everything’s more dynamic,” he said. “And you need to be adaptive, you need to be ready to manage dynamic infrastructure, multi-region, multi-cloud, everywhere. You need to be able to monitor this change.”
Along with Product Manager Karishma Irani, Alberto delivered a presentation titled Infrastructure: Now You’re Playing With Power. As Karishma explained, since New Relic Infrastructure launched in November 2016, it has been adopted by more than 3,000 customers, making it the fastest growing product in New Relic history. In that time, the Infrastructure team has rolled out some 50 new features.
Alberto and Karishma added to the list by announcing several new Infrastructure integrations for AWS (Application Load Balancer, API Gateway, Elastic File System, BBC Flowlogs) and Microsoft Azure (VM, Cosmos DB). They also showcased improved alerting capabilities and enhanced container support. In fact, as Karishma explained, you can now run New Relic Infrastructure itself as a container image, which enables monitoring on even security-restricted systems.
A key takeaway, Alberto said, is that New Relic Infrastructure is investing heavily in containers and orchestration support, and continues to support a multi-cloud strategy.
Watch the full presentation here:
Building for a multi-cloud world
Even as AWS leads in these early days of the cloud race, other providers are promoting the value of a multi-cloud future. In his talk titled Open, Multi-Cloud Observability and Reliability at Planet-Scale, Google Engineering Director Jeromy Carriere emphasized Google’s commitment to an open, multi-cloud environment.
“What we’re trying to do is build tools that make Google Cloud a natural part of a multi-cloud, hybrid-cloud ecosystem,” he said. As an example, Jeromy cited Kubernetes, Google’s open-source system for automating deployment, scaling, and managing containerized applications, which was designed by Google and then donated to the Cloud Native Computing Foundation. “We have put an enormous amount of energy behind Kubernetes and we think that it is really dramatically changing the container ecosystem.”
In the same vein, Jeromy mentioned Spinnaker, a Continuous Integration/Continuous Deployment platform tool for cloud development, and its cross-cloud API platform Apigee. “We’re trying to build for a world that we know will be multi-cloud.”
Watch the full Google Cloud presentation here:
Making cloud count at scale
One company that has fully committed to the cloud—and never looked back—is Braze. Formerly known as Appboy, the Braze customer engagement platform helps brands deliver great customer experiences in today’s mobile-first world. And it does so on a massive scale.
“We have hundreds of millions of monthly active users and we’re sending them tens of billion of messages every month,” explained Braze co-founder and CTO Jon Hyman in his talk Operating in the Cloud for 850 Million Monthly Active Users. Those message include emails and push notifications for brands like Domino’s Pizza, SoundCloud, and Lyft. Operating at such a scale simply wouldn’t be possible, Jon said, without cloud-driven flexibility.
That flexibility takes the form of auto-scaling, which helps Braze add and drop resources as needed. How? By isolating clusters, availability zones, and databases in real time.
“A cluster has API servers that ingest data,” Jon explained. “It has servers that are out processing that data and sending messages. It even has different web dashboards for customers to log into and interact with. And so we can actually create different clusters to scale essentially horizontally at a very high level.”
“Auto-scaling is challenging,” Jon conceded. “It requires knowing what’s going to happen in the future, being able to anticipate it. And it requires knowing what’s happening now in order to do correctly. Through monitoring and analytics, you can actually get that visibility.”
That’s where New Relic comes in. By continuously gathering and interpreting performance data, New Relic helps Braze run a script every five minutes on AWS that either spins up or scales down servers.
To learn more about how New Relic helps Braze with cloud-related incident handling and bug fixing at scale, watch Jon’s full talk here:
FutureStack: New York photos by Andres Otero.