Google Cloud’s partnership with Sky Group, one of Europe’s largest media and entertainment companies, dates back more than four years to when Sky first became a Google Cloud customer moving diagnostic data from millions of its Sky Q TV boxes to its Google Cloud data platform.

In June 2019, a few years into their cloud adoption journey, Sky was faced with a challenge they had anticipated from the start. Their recent bill across all major cloud providers had been increasing rapidly, reaching their planned yearly budget after only six months. Sky wasn’t sure if they’d undershot their forecasts, if they were overspending, or both.

“In the beginning, we were given a brief to investigate internal cloud spend with the aim of finding out where we could make savings, but in reality we didn’t know what we would expect to find,” said Nathan King, a cloud architect in the Cloud Enablement Center and now Head of Cloud Financial Management (FinOps) at Sky since the start of 2020.

Nathan assembled a small team who started to explore Google Cloud spend using the Cloud Billing tool. At first, they drilled into their biggest Google Cloud cost categories and discovered some immediate cost optimizations with BigQuery, Compute Engine and Cloud Storage. Over the course of the next six months, through careful analysis, they managed to find over $1.5m in immediate savings, exceeding expectations.

Yet they soon realized this was just the tip of the iceberg—it was clear there were millions of pounds more savings to be made, but actually achieving them at scale would require careful planning. “We formed a FinOps function to target these savings, but with 600 to 700 projects for Google Cloud alone, spanning four Google Cloud organizations, it would have been a manual process and difficult for teams to digest our recommendations,” Nathan said.

After attending a Google-led FinOps workshop and shaping their FinOps strategy, Nathan’s team focused on iterating through the FinOps lifecycle phases of Inform, Optimize, Operate and generating savings over time. Here’s how they did it:

Inform: Make Information Visible

The first step was focused on developing a clear vision for cost allocation and recharge, which required partnering closely with the finance, procurement and tax teams (particularly for international and affiliates) to understand the supporting business logic and processes. With a lot of hard work, the team managed to break down barriers to implement and embed new processes into broader business functions like finance.

WIth the recharge model in place, the team ran a number of pilots to find the right FinOps tooling to meet their needs. They ran a number of pilots, including using Data Studio and visualizing BigQuery exports. Given their ambitions to scale across the enterprise globally, the team chose Google Cloud’s Looker to realize their vision, building intuitive dashboards to visualize spend and recommendations across all cloud providers. “We wanted one view across all clouds, where customers can dynamically see cloud spend and intelligent optimization recommendations in just one place,” Nathan said.

After less than three weeks of development, the Looker dashboards were ready to go and have been a game changer ever since. “The moment our leadership and different departments started seeing the Looker dashboards, the value we were adding as a FinOps team became immediately clear,” Nathan said.

There are different report pages for each stakeholder group, each custom developed and automated using Looker and BigQuery. The BigQuery Optimization page, for example, provides insights on Slots consumed across the organization, down to granular query data like the cost of each query, how it was written, who submitted it and number of slots utilized. The dashboards also highlight potential areas of optimization, like BigQuery datasets without retention policies set or where data isn’t partitioned.

A recent breakthrough has been building pages for business teams, showing the related cloud spend contributing to a business unit of value, such as the cost per live stream or per subscriber in Sky’s case. Although this is an inherently difficult metric to capture, the opportunity has been made possible with the FinOps team’s progress and is starting to drive business investment decisions.

Optimize: Drive Cloud Efficiency

The second stage of the FinOps lifecycle focuses on delivering optimizations. As Sky’s FinOps dashboards were operationalized and highlighted savings opportunities, they enabled users to generate more than $3 million in Google Cloud savings alone in 2020 and over $800,000 in other cloud providers.

The team began with focusing on the top four products by spend: BigQuery, Compute Engine, Cloud Dataflow and Cloud Storage. Working with their Google account team and studying Google whitepapers and blog posts like Cloud cost optimization: principles for lasting success, they developed their own best practice guidance and embedded recommendations into the dashboards.

Creating their own recommenders and leveraging Google Cloud’s recommenders, the team discovered a plethora of cost optimization opportunities. “Key examples were overly expensive queries, storage buckets set without retention policies, and VMs without autoscaling enabled,” Nathan said. Teams were then empowered to make their own savings, like the NowTV business unit that had been forecast to overspend for the year until they received their dashboard with thousands of optimization recommendations. After just three weeks, the team had implemented more than 90% of recommendations and brought their spend under budget for the year, saving more than 50%.

The FinOps team still searches for new recommendations every day and have been collaborating with Google product managers to take their insights to the next level. “We’ve loved partnering with Google product managers, who encourage us to give feedback on new features before they go to market. We’ve also shared some of our in-house recommenders to influence the features being developed by Google, including the Idle VM and Idle Persistent Disk Recommenders as part of Active Assist,” Nathan said.

Operate: Embed FinOps & Drive Self-Sufficiency

Now that teams could visualize their cloud spend and make real-time decisions based on cost optimization recommendations, the FinOps team has begun working on embedding processes, leveraging machine learning, and improving efficiency in their own ways of working.

Looker’s extensive capabilities continue to play a role in this. “Before we started using Looker, our most popular report was an electricity bill showing customers’ detailed monthly cloud spend, previous month comparisons and forecasts for months ahead,” Nathan said. “This report took days, sometimes weeks to run. With Looker, we’ve automated the entire process and brought that time down to just minutes.”

More teams are embedding the dashboards into their own processes, like finance, which now uses the interactive dashboards in meetings instead of static report snapshots, or in-house Google Cloud architects, who use the recommendations to optimize their cloud spend before deploying any technology.

As the FinOps team continues to operate like a product function, designing with CX/UX in mind and iteratively releasing new features like anomaly reporting, budget alerts, and forecasting based on machine learning, it’s becoming clear that Cloud Financial Management is a key capability and mindset that can impact wide-reaching parts of the business at scale.

Elevating Sky’s FinOps journey to the next level
Indeed, as more business teams collaborate with the FinOps function, the opportunities are growing. “The FinOps team has changed the way we view and manage cloud spend, enabling us to partner with finance and show digestible reports to the CFO. We’re now looking further to broaden our range of insights, like elevating our dashboards to understand how using Google Cloud is supporting Sky’s Net carbon zero ambitions by incorporating Google’s data center sustainability metrics,” says Vince Marco, Architecture Manager at Sky.

So, after being unsure of drivers for their increasing cloud spend in 2019, 18 months later Sky is far more confident about its investment decisions. The team knows that every dollar spent is being used optimally and driving maximum value for its investment.

If you’re an enterprise using cloud, but want to better manage cloud costs, consider setting up a FinOps capability and creating a FinOps mindset. Looker can help you get started by providing reporting and insights into cloud expenditures to identify initial savings. As you learn more and scale, empower teams to make their own savings utilizing built-in actionality for monitoring and customizing for business billing activity nuances and department-specific chargebacks. Reimagine how cloud finances can be managed and optimized as Sky is doing.

To learn more about Looker’s Cloud Cost Management Block visit Looker Marketplace.

Compute Engine lets you create and run virtual machines (VMs) on Google’s infrastructure, allowing you to launch large compute clusters with ease. When it comes to getting started with Compute Engine, our customers have lots of questions—but some questions come up more often than others. 

We looked at an internal list of the most popular Compute Engine documentation pages over a 30-day period to find out what topics were explored by users again and again. Here are the top four questions users have about Compute Engine, in order.

1. What are the different machine families

Compute Engine lets you select the right machine for your needs. You can choose from a curated set of predefined virtual machine (VM) configurations optimized for specific workloads, ranging from small-level purpose to large-scale use cases or create a machine type customized to your needs with our custom machine type feature. 

Compute Engine machines are categorized by machine family, including: 

• General-purpose: Best price-performance ratio for a variety of standard and cloud-native workloads 
• Compute-optimized: Highest performance per core for compute-intensive workloads, such as ad serving or media transcoding  
• Memory-optimized: More compute and memory per core than any other family for memory-intensive workloads, such as SAP HANA or in-memory data analytics 
• Accelerator-optimized: Designed for your most demanding workloads, such as machine learning (ML) or high performance computing (HPC)

Read the documentation to learn more about each machine family category.

2. How to connect to VMs using advanced methods

In general, we recommend using the Google Cloud Console and the gcloud command-line tool to connect to Linux VM instances. However, some of our customers want to use third-party tools, or require alternative connection configurations. 

In these cases, there are several methods that might fit your needs better than the standard connection options:

• Connecting to instances using third-party tools (e.g. Windows PuTTY, Chrome OS Secure Shell app), or MacOS or Linux local terminal) 
• Connecting to instances without external IP addresses
• Connecting to instances as the root user Manually connecting between instances and running commands as a service account

Read the documentation to learn about advanced methods for connecting Linux VMs.

3. How to set up OS Login

OS Login lets you use IAM roles and permissions to manage access and permissions to VMs. 

OS Login is the recommended way to manage users across multiple instances or projects. OS Login provides:

• Automatic Linux account lifecycle management
• Fine-grained authorization using Google IAM without having to grant broader privileges
• Automatic permissions updates to prevent unwanted access
• Ability to import existing Linux accounts from Active Directory (AD) and Lightweight Directory Access Protocol (LDAP)

You can also add an extra layer of security by setting up OS Login with two-factor authentication or manage organization access by setting up organization policies.

Read the documentation to learn how to configure OS login and connect to your instances.

4. How to manage SSH keys in metadata 

Compute Engine allows you to manually manage SSH keys and local user accounts by editing public SSH key metadata.

You can add  public SSH keys to instance and project metadata using: 

• The Google Cloud Console The gcloud command-line tool 
• API methods from the Google Cloud Client Libraries

Read the documentation to learn how to manually manage SSH keys and local user accounts in metadata.

Don’t see your question here? Check out the Compute Engine documentation for all of our recommended guides, tutorials, and resources.

The typical smart factory is said to produce around 5 petabytes of data per week. That’s equivalent to 5 million gigabytes, or roughly 20,000 smartphones.

Managing such vast amounts of data in one facility, let alone a global organization, would be challenging enough. Doing so on the factory floor, in near-real-time, to drive insights, enhancements, and particularly safety, is a big dream for leading manufacturers. And for many, it’s becoming a reality, thanks to the possibilities unlocked with edge computing.

Edge computing brings computation, connectivity, and data closer to where the information is generated, enabling better data control, faster insights, and actions. Taking advantage of edge computing requires the hardware and software to collect, process, and analyze data locally to enable better decisions and improve operations.

At Hannover Messe 2022, Intel and Google Cloud will demonstrate a new technology implementation that combines the latest generation of Intel processors with Google Cloud’s data and AI expertise to optimize production operations from edge to cloud. This proof-of-concept project is powered by the Edge Insights for Industrial platform (EII), an industry-specific platform from Intel; and a pair of Google Cloud solutions: Anthos, Google Cloud’s managed applications platform, and the newly-launched Manufacturing Data Engine.

Edge computing exploits the untapped gold mine of data sitting on-site and is expected to grow rapidly. The Linux Foundation’s “2021 State of the Edge” predicts that by 2025, edge-related devices will produce roughly 90 zettabytes of data. Edge computing can help provide greater data privacy and security, and can accomodate the reduced bandwidth needs between local storage and the cloud.

Imagine a world in which the power of big data and AI-driven data analytics is available at the point where the data is gathered to inform, make, and implement decisions in near real-time.

This could be anywhere on the factory floor, from a welding station to a painting operation or more. Data would be collected by monitoring robotic welders, for example, and analyzed by industrial PCs (IPCs) located at the factory edge. These edge IPCs would detect when the welders are starting to go off spec, predicting increased defect rates even before they appear, and adding preventive maintenance to correct the errors without any direct intervention. Real time, predictive analytics using AI could substantially prevent defects before they happen. Or the same IPCs could use digital cameras for visual inspection to monitor and identify defects in real-time, allowing them to be addressed quickly.

Edge computing has powerful potential applications in assisting with data gathering, processing, storage and analysis in many manufacturing sectors, including automotive, semiconductor and electronics manufacturing, and consumer packaged goods. Whether modeling and analysis is done and stored locally or in the cloud, or is predictive, simultaneous, or lagged, technology providers are aligning to meet these needs. This is the new world of edge computing.

The joint Intel and Google Cloud proof of concept aims to extend the Google Cloud capabilities and solutions to the edge. Intel’s full breadth of industrial solutions, hardware and software, are coming together in this edge-ready solution, encompassing Google Cloud industry-leading tools. The concept shortens the time to insights, streamlining data analytics and AI at the edge.

The Intel-Google Cloud proof of concept demonstrates how manufacturers can gather and analyze data from over 250 factory devices using Manufacturing Connect from Google Cloud, providing a powerful platform to run data ingestion and AI analytics at the edge.

In this demonstration in Hannover, Intel and Google Cloud show how manufacturers can capture time-series data from robotic welders to inspect welding quality and show how predictive analytics can benefit the factory operators. In addition, the video and image data is captured from a factory camera to show how visual inspection can highlight anomalies on plastic chips with model scoring. The demo also features zero-touch device onboarding using FIDO Device Onboard (FDO) to illustrate the ease with which additional computers could be added to the existing Anthos cluster.

By combining Google Cloud’s expertise in data, AI/ML and Intel’s Edge Insight’s for Industrial platform that was optimized to run on Google Anthos, manufacturers can run and manage their containerized applications at the edge, in on-premise data center, or in public clouds using an efficient and secure connection to the Manufacturing Data Engine from Google Cloud. It forges a complete edge-to-cloud solution.

Simplified device onboarding is available using Fido Device Onboard (FDO)—an open IoT protocol that brings fast, secure, and scalable zero-touch onboarding of new IoT devices to the edge. FDO allows factories to easily deploy automation and intelligence in their environment without introducing complexity into their OT infrastructure.

The Intel-Google Cloud implementation can analyze that data using localized Intel or third-party AI and machine learning algorithms. Applications can be layered on the Intel hardware and Anthos ecosystem, allowing customized data monitoring and ingestion, data management and storage, modeling, and analytics. This joint PoC facilitates and support improved decision making and operations, whether automated or triggered by the engineers on the front lines.

Intel collaborates with a vibrant ecosystem of leading hardware partners to develop solutions for the industrial market by using the latest generation of Intel processors. These processors can run data intensive workloads at the edge with ease.

Putting data and AI directly into the hands of manufacturing engineers can improve quality inspection loops, customer satisfaction, and ultimately the bottom line.

The new manufacturing solutions will be demonstrated in person for the first time at Hannover Messe 2022, May 30–June 2, 2022. Visit us at Stand E68, Hall 004, or schedule a meeting for an onsite demonstration with our experts.