Google Cloud Technology Nuggets — December 1–15, 2022 Edition

Welcome to the December 1–15, 2022 edition of Google Cloud Technology Nuggets.

Predictions on how IT will change

Prediction #1: By 2025, 90% of security operations workflows will be automated and managed as code.

Prediction #2: By 2025, 4 out of 5 enterprise developers will use some form of curated open source

These are some predictions that our IT leaders made at Google Cloud Next ’22 about how IT is likely to change in the coming years. Check them out to learn more: Prediction #1 and Prediction #2.

Application Modernization

API Management and Service Mesh, when used in combination can help an organization streamline how they expose their microservices to both internal and external users. At the same time, gaining visibility into their operations with full lifecycle management and performance analytics for their APIs. Check out this article that explains how this comes together.

Kubernetes and GKE

If you have cross-project backends, what would be the best way to expose these to the Internet, without the overhead of having multiple Google Cloud Global Load Balancer (GLBC). This guide presents a neat solution in the form of using a Multi-Cluster Ingress along with Anthos Service Mesh to achieve the same.

If you are managing clusters across your organization, how do you ensure that they have been following the best practices for clusters. Help is now available in the form of an open source tool called GKE Policy Automation, that deploys a solution on top of Google Cloud services. This tool audits your cluster across the organization, can be run one-off or at regular intervals and is integrated with the Security Command Center. Check out the blog post for more details.

Startups

Time is of the essence for startups. The teams want to iterate as quickly as possible. Serverless and fully-managed services go a long way in reducing the toil for developers and in architecting solutions. Check out this guide that helps you understand how Google Cloud helps startups to iterate quickly. The article covers where you can run your code, our IDE extension Cloud Code that helps the inner loop of development, services to help you orchestrate between your own services and even where you can use your own custom/batch code.

Storage, Databases and Data Analytics

AlloyDB for PostgreSQL, a fully-managed, PostgreSQL-compatible database service is now Generally Available. It has full PostgreSQL compatibility with the best of Google: scale-out compute and storage, integrated analytics, and AI/ML-powered management.

Since the preview announcement in May this year, several features have been added. These include: security features like customer-managed encryption keys (CMEK) and VPC Service Controls, a preview of cross-region replication and additional configuration options. Check out the blog post for more details on learning more about AlloyDB, customer momentum and more.

If you have been managing SQL databases in the cloud on your own, we have some data from our customers directly on their experiences in using Cloud SQL, our fully-managed SQL databases offering. The whitepaper captures the inputs provided by the customer directly.

If you are a Cloud Pub/Sub user, one of the challenges that you might be facing is that you have to handle your own processing logic to avoid duplicate messages. This is now tackled via the exactly-once delivery feature that is now generally available in Cloud Pub/Sub. Check out the post that provides an interesting writeup on the benefits of this feature and how it has been implemented.

Continuing on the Pub/Sub theme, the Pub/Sub Group Kafka Connector is now generally available. This will allow you to keep your Google Cloud systems and Kafka based systems on-premise in sync. Both source and sink connection options are available in the Connector. Check out the post for more details.

What is the best file format to use when it comes to importing large datasets into BigQuery? Should these file formats be compressed or uncompressed? Check out this blog post that does a shootout between multiple file formats (CSV, AVRO, PARQUET, etc) and presents which one is the most efficient.

Did you hear about the features through the year that made BigQuery SQL more user friendly? I am sure there is a feature or two that will surprise you in this blog post.

Machine Learning

Google Translation Hub that was released earlier this year has been seeing some good traction from customers. Check out this blog post that helps you understand the various features of Google Translation Hub, especially for enterprises, and how they are making use of them,

The process of evaluating the quality of a ML model continuously can turn out to be a difficult and costly task. Vertex AI Model Evaluation enables you to iteratively assess and compare model performance at scale. With Vertex AI Model Evaluation, you define a test dataset, a model, and an evaluation configuration as inputs and it will return model performance metrics whether you are training your model using your notebook, running a training job, or an ML pipeline on Vertex AI. Check out this blog post for more information.

Continuing on a similar thread, if you are looking to improve the process of training PyTorch based deep learning models via distributed training methods, check out this post.

APIs

How do you design your APIs to be readable, less chatty, flexible and more. Check out 6 common mistakes to avoid in RESTful web API Design.

API security is paramount for any organization. With public APIs constantly under threat, how do you design an API security strategy? Check out this post.

Developers and Practitioners

If you have been using Google Cloud for a while, you have probably dealt with Billing Alerts, where you can get notified when the charges for the month cross certain thresholds. How about looking at controlling your costs by automating the process of receiving these budget alerts and then taking some action (say stop a VM?). Check out this article that deploys a DeployStack solution named Cost Sentry that sets up a solution for you using Google Cloud services that you can customize further to implement the actions that you would like to take when budgets cross certain thresholds.

Cloud Run currently has a Preview feature for running jobs (Cloud Run Jobs). A job is a long running process that runs its tasks and exits when finished. It does not listen for or serve requests. What kind of functionality could you run in a job? Well, A database migration or export could be one such task that you could abstract away inside a Cloud Run job. Check out this article that covers running database migrations with Cloud Run jobs.

Check out the final part of the series on Workflow Patterns and Best Practices, where the authors cover workflow lifecycles and benefits of using Firestore within Workflows. Check our Part 1 and Part 2 of the series too.

If you are looking for an interesting application to build out this month, how about predicting a movie score and in the process learning about Vertex AI, BigQuery and MongoDB Atlas? Check out this article that gives a step by step process on how to build this application.

Learn about Google Cloud

December is a time to get a good break and start afresh in the new year. If you are taking some time off and Google Cloud is on your radar to learn, try out multiple avenues available to you to learn more. These range from hands-on labs, webinars, competitions and more. Check out the guide that provides 12 no-cost ways to learn Google Cloud over the holidays.

Apigee is Google Cloud’s natively available API Management service. While you might be familiar with general API terms, Apigee at times, has its own definition of things. Check out this guide that helps you demystify commonly used Apigee terminology,

Stay in Touch

Have questions, comments, or other feedback on this newsletter? Please send Feedback.

Looking to keep a tab on new Google Cloud product announcements? We have a handy page that you should bookmark → What’s new with Google Cloud.