Simon Poulton, CEO, Kasna
A few weeks ago the Kasna team journeyed to the hills of San Francisco for Google Cloud Next ‘19. We’ve got some detailed blogs in the pipeline on some interesting topics but we wanted to catch you up on some of the major themes and give them a local perspective. We’re also running a recap event where we’ll drill down into much more detail than what’s been covered here, so if you work in a local enterprise or startup and want to join us for breakfast (with great food and St Ali barista made coffee) on Wednesday 8 May, please register here.
The first thing that seems to have most people’s chins fiercely wagging is definitely Anthos, Google Cloud’s hybrid-cloud ecosystem product. Anthos is an evolution of the previously announced product, Cloud Services Platform, but it was the addition of some soon to arrive features which really made the news. Anthos will allow management of kubernetes applications on-prem, in GCP, AWS & Azure. While I’ll leave a more in-depth tech analysis of this to my compadre @Troy Bebee in a later post, this is hugely interesting from a local perspective. Many of the customers we speak to, particularly the large ones, have more than one cloud vendor in play, so the idea of being able to manage containers using a single pane of glass across a complex application estate is certainly appealing. It’s an opportunity to accelerate, scale and control the enterprise journey to the cloud across all platforms. Adding migration and Config Management tools to the mix also reminds me of Microsoft Exchange vs Lotus Notes 20 years ago (for those of you grey enough to remember), when Microsoft created interoperability and migration tools for Notes, which essentially helped them win many enterprises to their Mail cause all those years ago. Not that I’m comparing our friends at Mountain View to the guys at Redmond, nor the other providers to Lotus!!
Global resiliency for your services
Secondly, Google have made a further serious play for the non-cloud-native enterprise workloads. SAP was mentioned many times, new VM limits of up to 12TB of memory (available later this year), BYOL (Bring your own licence), and Shielded VMs are all features which will make running an enterprise VM in GCP a more attractive offer. The further announcements around managed services for Active Directory and MS SQL show that Google has committed to both running the supporting infrastructure for enterprise VMs and reducing the cost overhead for those platforms by leaving the non-differentiating management work to the scale of Google.
On the networking front, you can see the ‘top end of town’ play with 100Gpbs Cloud Interconnect won’t be something many Australian companies will be ordering next week, but the freshly announced High Availability VPN is likely to be something compelling for many given its 99.99% SLA (when it comes out of beta). Google also added Traffic Director to ease the burden of multi-regional deployments using a service mesh. Care to make your application more resilient without doing the hard work?
First of its kind capability
That brings us to everyone’s favourite (and probably the prickliest) topic in cloud, Security. As we’ve all experienced on Cloud projects… security can be the thing that takes the longest, and the thing which brings us undone – so what have Google announced here of relevance?
A beta of a user interface for Google’s DLP (Data Loss Prevention) is a good place to start, meaning you no longer need to use the API to leverage DLP in GCP. From an Australian perspective, where everything from Tax File Numbers to drivers licence formats are unique, not just in-country, but in-state, it’s now going to be far easier for us to unite the knowledge of what sensitive data looks like with the act of finding and (where appropriate) de-identifying this type of information in the cloud. The GA of several other tools, including VPC Service Controls and Cloud Security Command Center mean we can define and manage the security perimeter, as well as manage both the risk and occurence of events pertinent to security professionals.
A fully managed serverless product
As a DevOps pundit of around eight years I was keen to see what Google would announce on the developer experience front, and they didn’t disappoint with some great new stuff. Cloud Run is essentially a serverless capability without the usual limitations requiring a developer to choose a supported language. Based on Knative it also allows you to choose Cloud Run, or run the code on your own kube cluster. In Australia this should increase the number of teams able to quickly leverage serverless using their existing code, unlocking loads of time, effort and cash previously spent managing infrastructure to be redirected to valuable in app features. Cloud Code also gained some attention as a way to focus on code rather than config, by abstracting some of the kube/cluster work away from the devs. Again the benefit here is more time on features, less on infra. VS Code was in beta and IntelliJ in alpha at the time of the announcement.
There were loads of juicy tidbits to choose from in the Data, AI & ML segment. Data Fusion was announced in beta, as a data integration service for the hybrid data enterprise. The main change in approach here is to target the GUI based tool at people without extensive Data Engineering and/or coding experience, broadening the number of people in our talent-constrained market here who can now ingest, transform and get insight from valuable data. Another enhancement in this vein was Connected Sheets for Big Query, which enables people familiar with Google or Excel spreadsheets to manipulate huge reams of data natively in Google Sheets. This should allow statistics and numbers savvy users like Finance or Marketing departments to leverage BigQuery data sets without having to engage specialist folks in the Data team.
Google continued to enhance the AI & ML platform in both features and ease of use. From letting teams manage ML projects via AI Platform, to new AI services for Recommendation and Document understanding, we’ll explore a set of these at the above mentioned event along with some key takeaways from TensorFlow world.
You can get a full list of the announcements direct from Google, here.
All the links a bit much reading for you? Join us on Wednesday 8 May and we’ll take you through it face to face – it’s our mission to make Google a more accessible and valuable platform for you and your business at scale!