Last week in San Francisco was the Google Cloud Next ‘17 conference. Over ten thousand attendees came through the Moscone Center over three days, and Google announced dozens of products, partnerships, and updated features for Google Cloud. As we held down the Scalr booth on the show floor, we spent those days talking with customers and hearing what developers, engineers, and admins thought of the keynotes.
Here’s the top three things everyone mentioned:
Cloud Spanner. Even though it was released in General Availability earlier in the month, the Day 2 Keynote brought Cloud Spanner to everyone’s attention. Cloud Spanner enables you to create horizontally scalable, globally consistent relational databases. This is powerful for enterprises and large companies, when quick availability of databases across regions is problematic as you process millions of queries a day across apps and services. Now you can scale easily and maintain consistency. Google has been using Cloud Spanner internally for years to handle millions of queries per second, so it’s been through the paces at scale.
Increased Portability and Freedom. While Google Cloud Platform isn’t a vendor lock-in, a common complaint on the show floor was that it wasn’t easy to get applications out of GCP unless they are containerized. In response, Google has made it easier to bring apps in and out of GCP with AppEngine Flex. In addition, AppEngine now has support for every major development language now: Node.js, Ruby, Java 8, Python 2.7 or 3.5, Go 1.8, plus PHP 7.1 and .NET Core in beta. Many people, including myself, were excited with how much Google Cloud Platform has evolved so much over the last year. Once the price wars between AWS and GCP were mostly over, now Google could focus on becoming the infrastructure for web applications, mobile backends, and APIs.
Focus on Technology, Not Business Partnerships. Many attendees didn’t enjoy the Day 1 Keynote simply because it was companies like Disney, SAP, Verizon, HSBC, Colgate-Palmolive, eBay, and The Home Depot focusing on how deeply nested they were with Google infrastructure. SAP announced that SAP HANA would be available on GCP. Eric Schmidt, CEO of Alphabet, declared that they poured thirty billion dollars into GCP and jokingly implored companies not to waste time building out their own infrastructure. Admittedly it was all necessary to demonstrate the popularity of GCP with the enterprise crowd and how well it works with the rest of Google’s products like G Suite.
However, developers and admins alike felt that it would have been important to jump right into the latest product updates, like the 8% reduction in price for Compute Engine, or that Engineering Support can pair your company engineers with Google engineers to help diagnose infrastructure problems.
The energy was interesting at Next ‘17. Many attendees we talked to had segments of their infrastructure in GCP (mostly Kubernetes on GKE), but there were also a good number of startups there that were going GCP first. It felt like many just curious about what Google had to offer. And of course, there was a VR playspace for everyone to decompress at.
The usual booths were there - Ubuntu, Redhat, Scalr - but there were also companies that were ‘Google recommended’ tools for G Suite (Like Prosperworks). It felt odd, but we got to see the enterprise from both sides - the product/application powerhouse of GCP and the logical/business unit side of G Suite. With Amazon coming at them on the work side with Chime, WorkDocs, and WorkMail, it makes sense to show off the scope of what Google is offering.
Here’s the highlights of the announcements from Google Cloud Next 2017.
Google Cloud Functions - While it’s not fair to parallel everything to existing Amazon services, Google Cloud Functions is similar to AWS Lambda. Cloud Functions are a way to create microservices without spinning up servers that just handle webhooks. Cloud Functions listens for events emitted by cloud services, then runs your single-use code snippets in response.
Titan - Titan is a Google-designed security chip to be installed inside Google data centers, that can identify and authenticate access at the hardware level. It is built to ensure security for the enterprises at the hardware level. Cloud may be considered unstable because of pricing, but Titan is a way to show that even though cloud is already secure, this takes it a step further. Titan also includes a random number generator, performs cryptographic operations in the isolated memory, and has a dedicated on-chip secure processor. It’s so small, Urs Hölzle announced it while wearing it on his ear piercing.
BigQuery Data Transfer Service Still in private beta, but Transfer Service helps companies get value from all their Google-managed advertising datasets. Marketing analysts can import data from Google Adwords, DoubleClick Campaign Manager, DoubleClick for Publishers and YouTube Content and Channel Owner reports. While this is great for companies that focus on diverse ad and media content, this is part of a new evolution of GCP where it’s becoming easier to bring in commercial datasets alongside internal data sets to leverage your business intelligence.
Cloud Container Builder - Cloud Container Builder is standalone tool that lets you build your Docker containers on GCP regardless of your deployment environment. It’s a fast way to package applications into containers as part of your workflow.
Cloud Storage for Firebase - Firebase Storage is now Google Cloud Storage for Firebase and adds support for multiple buckets, support for linking to existing buckets, and integrates with Google Cloud Functions. If you’ve never used Firebase as the backend for your mobile applications, they were acquired by Google last year to expand on their mobile infrastructure support. Firebase backend suites give you analytics, storage, auto-scaling databases and infrastructure, and a pretty straightforward toolset to get applications live.
12-month Free Trial with $300 credits - Google has gone all in and extended their GCP free trial from 60 days to 12 months. With this, you can use your $300 credit across all GCP services and APIs. Of course, if you have more than a couple services active you’ll hit that $300 limit in a few months, but it’s nice to have the ability to slowly test services (like the ease of deploying Docker containers) across cloud providers and see what works best.
Always Free - Similar to the Free Tier for instances in AWS EC2. Always Free non-expiring usage limits that you can use to test and develop applications at no cost. This only works if you have never used the free tier with your Google account. (our your company Google account).
Cloud Machine Learning Engine (GA) - This is for organizations that want to train and deploy their own models into production in the cloud. Google had a Codelabs area at the conference where they had engineers giving demos and tutorials, and a sit down area where you could work at your own pace.
It’s exciting to see how the cloud industry continues to mature. Now that the price wars are mostly settled and features are consistent across clouds, we can be excited for what happens in the next year. Machine learning continues to move from buzzword to in production clusters, and it’s interesting to see how the private cloud providers will respond. Will they focus on abstracting the boundaries between public and private clouds? Or join in on the feature wars? While some enterprises like Disney and The Home Depot are migrating massive sections of their infrastructure to the public cloud, there is still legacy infrastructure that is here to stay (even if only for the next decade).
Here’s the entire list of annoncements: https://blog.google/topics/google-cloud/100-announcements-google-cloud-next-17/