An Annotated History of Google’s Cloud Platform

Described without… help from marketing

See Also: What are the Google Cloud Platform services?

Google Cloud Platform (GCP) has been growing rapidly as Google invests heavily in it development. Sadly, the Wikipedia entry for GCP is garbage, and it’s difficult to skim the blog as a timeline. For my own sake, I wrote this attempt at an objective timeline of significant events and launches for GCP.

• Pre 2008 — Computers invented. Google Founded.
• April 2008 — App Engine launches in preview for 20,000 developers (10,000 at launch, then expanded by another 10,000 four days later), as a tool to run web applications on Google’s infrastructure. Applications had to be written in Python, and were limited to 500MB of storage, 200M megacycles of CPU per day, and 10GB bandwidth per day. Services offered included dynamic webserving, persistent storage (Datastore), and APIs for authenticating users and sending email¹.
• May 2008 — The App Engine preview opens signups to all developers. Two new APIs are also announced: An image manipulation API and Memcache².
• February 2009 — App Engine allows developers to purchase additional computing resources beyond the free tier quotas³.
• April 2009 — App Engine announces the availability of support for the Java programming language⁴.
• May 2010 — Cloud Storage launched. Google App Engine for Business announced, offering added management and support features tailored specifically for the enterprise. BigQuery and Prediction API announced in limited preview⁶.
• October 2011 — Cloud SQL is announced in preview offering a fully-managed cloud-based relational database service for App Engine⁷.
• November 2011 — App Engine officially leaves preview and becomes a “fully supported Google product”⁸.
• March 2012 — A stable version of the Go programming language (Go 1), libraries, and tools is released, along with a (still experimental) App Engine SDK for the Go runtime.
• April 2012 — Google announces a change in their deprecation policy for all developer products; App Engine’s deprecation period is set to one year⁹. BigQuery, first presented in 2010, goes into General Availability (GA)¹⁰.
• June 2012 — App Engine turns up its first non-US cluster, based in the European Union for a limited number of customers¹¹. Compute Engine is launched into preview¹².
• July 2012 — Launch of the Google Cloud Platform Partner Program¹³.
• October 2012 — App Engine experiences a major outage that also affected Tumblr and Dropbox¹⁴.
• February 2013 — Launch of paid support packages for services on GCP, including options for 24x7 phone support and consultation, and direct access to a Technical Account Manager (TAM)¹⁵.
• May 2013 — After a 12-month preview, Compute Engine is released to GA¹⁶. App Engine for PHP is announced and made available in Limited Preview.
• August 2013 — Cloud Storage begins automatically encrypting each Storage object’s data and metadata under the 128-bit Advanced Encryption Standard (AES-128), and each encryption key is itself encrypted with a regularly rotated set of master keys¹⁷.
• October 2013 — PHP support for App Engine goes into Preview, making PHP on App Engine available for everyone¹⁸.
• November 2013 — After 18 months, Cloud Endpoints graduates to GA, along with the associated Mobile Backend Starter kit.
• February 2014 — Cloud SQL graduates to GA, supporting automatic encryption, 99.95% uptime SLA, and support for databases up to 500GB¹⁹. App Engine Modules also become Generally Available.
• March 2014 — APIs Client Library for .NET graduates to GA²⁰.
• March 2014 — A price drop affecting all products of between 30% and 85% is announced. Managed Virtual Machines and BigQuery Streaming are announced, the later allowing ingestion of 100,000 records per second per table²¹,²².
• April 2014 — Compute Engine zones, and associated support for deploying Cloud Storage and Cloud SQL, open in Asia Pacific²³.
• May 2014 — Stackdriver is acquired and becomes part of GCP. Cloud Storage JSON API goes into GA.
• June 2014 — Kubernetes, an open source container manager, is launched, along with improved support for Docker images²⁴.
• June 2014 — HTTP Load Balancing and SSD-based Persistent Disk are announced.
• June 2014 — Cloud Dataflow is announced in Limited Preview²⁵.
• July 2014 — Microsoft, RedHat, IBM, Docker, Mesosphere, CoreOS and SaltStack join the Kubernetes community and commit to actively contributing²⁶.
• August 2014 — SSD Persistent Disk becomes GA.
• September 2014 — gcloud-node, as idiomatic library for Node.js is launched with support for Cloud Datastore and Cloud Storage²⁷.
• October 2014 — Firebase is acquired and becomes part of GCP.
• October 2014 — GCP commits to standardized naming for release process: Beginning with “Alpha” for testing by a select group of customers, followed by “Beta” which is publicly available to to all customers, and finally “General Availability” at which point they gain an SLA and full support²⁸.
• November 2014 — Alpha release of Container Engine, a Kubernetes-powered, fully-managed cluster manager for Docker containers²⁹. Managed VMs, Local SSDs, and Compute Engine Autoscaler go into Beta.
• January 2015 — Google Cloud Monitoring (part of Stackdriver) goes into Beta with support for App Engine, Compute Engine, Cloud Pub/Sub, and Cloud SQL³⁰. Google Cloud Trace and Container Registry also go into Beta. Local SSD becomes Generally Available.
• March 2015 — Cloud Pub/Sub goes into Beta, Cloud Storage Nearline launched in Alpha.
• April 2015 — Cloud DNS goes into GA along with expansion of load balancing solutions to 12 additional points of presence (Los Angeles, San Francisco, Chicago, Seattle, Dallas, Miami, London, Paris, Stockholm, Munich, Madrid, Lisbon). BigQuery supports storage of data in European zones, and Data Flow goes to Beta.
• May 2015 — Bigtable is launched in Beta³¹.
• May 2015 — Launch of Compute Engine Preemptible VMs in Beta³².
• June 2015 — App Engine for PHP goes GA³³. Container Engine goes into Beta³⁴.
• July 2015 — Windows Server on Compute Engine³⁵, App Engine for Go³⁶, and Cloud Storage Nearline all graduate to General Availability. Kubernetes 1.0 is announced and described as “production ready”³⁷. A Go client for Bigtable is launched³⁸.
• August 2015 — Google Cloud Dataflow / Cloud Pub/Sub³⁸ and Container Engine / Deployment Manager graduate to GA³⁹.
• September 2015 — Compute Engine Preemptible VMs, Autoscaler, and 32 core VMs graduate to GA⁴⁰.
• October 2015 — New us-east1 region in South Carolina available for Compute Engine, Cloud Storage, and Cloud SQL. Cloud Storage regional buckets graduate to GA, making Taiwan (asia-east1), Belgium (europe-west1), South Carolina (us-east1), and Iowa (us-central1) available for Standard, Durable Reduced Availability, and Nearline Storage.
• October 2015 — Cloud Security Scanner graduates to GA⁴¹. Datalab and Cloud Shell launch in Beta⁴².
• November 2015 — Bebop is acquired and Diane Greene is hired to lead a new team combining all Google’s cloud businesses: Google for Work, Google Cloud Platform, and G Suite.
• December 2015 — Cloud Vision REST API launched in Alpha⁴³.
• February 2016 — Custom Machine Types⁴⁴ and Cloud Dataproc managed Hadoop and Spark service graduate to GA⁴⁵. Cloud Vision API enters Beta⁴⁶.
• March 2016 — Node.js on App Engine enters Beta⁴⁷. Cloud ML, Cloud Speech and Cloud Translate APIs launched in Alpha⁴⁸.
• March 2016 — Stackdriver launched into Beta, offering unified monitoring and logging for GCP and AWS⁴⁹.
• April 2016 — Cloud Vision API graduates to GA. Cloud CDN moves to Beta.
• May 2016 — Stackdriver Trace for App Engine graduates to GA. Ruby on App Engine goes Beta⁵⁰.
• July 2016 — Cloud Natural Language and Cloud Speech APIs enter open Beta⁵¹. Oregon Cloud Region (us-west1) is available, supporting Compute Engine, Cloud Storage, and Container Engine⁵².
• August 2016 — Cloud Shell, Cloud SQL, Bigtable and Datastore graduate to GA⁵³. Python runtime on App Engine Flexible Environment, with support for Python 3.4 and 2.7 goes into Beta⁵⁴.
• September 2016 — Apigee, a provider of application programming interface (API) management, acquisition announced.
• September 2016 — The locations of eight upcoming Google Cloud Regions — Mumbai, Singapore, Sydney, Northern Virginia, São Paulo, London, Finland and Frankfurt — are announced⁵⁵. Cloud ML goes Beta. New Google job role — Customer Reliability Engineer (CRE) — is announced.
• September 2016 — Stackdriver graduates to GA.
• October 2016 — Refresh of Cloud Storage with new storage classes (Coldline, Nearline, Regional, and Multi-Regional), data lifecycle management tools, improved availability, API compatibility across all storage classes and lower prices⁵⁶.
• October 2016 — Google Cloud announces it will join the .NET Foundation Technical Steering Group⁵⁷.
• October 2016 — Natural Language API graduates to GA⁵⁸.
• December 2016 — Idiomatic client libraries for BigQuery, Cloud Datastore, Stackdriver Logging, and Cloud Storage are available in Beta for C#, Go, Java, Node.js, PHP, Python, and Ruby.
• February 2017 — Cloud Endpoints and Google Cloud CDN become Generally Available, NVIDIA Tesla K80 GPUs for Compute Engine and Cloud ML enter public beta, and Intel Skylake processors are made available in Western / Eastern / Central US, Western Europe and Eastern Asia Pacific.
• February 2017 — Cloud Spanner, Google’s highly available, global SQL database is released into public Beta. It supports globally consistent reads across the entire database without locking. Supporting white paper describing its relationship with CAP theorem is also made available.
• March 2017 — Google Cloud Platform upgrades and migrates all issue tracking to a Issue Tracker hosted at issuetracker.google.com. Cloud Container Builder — a stand-alone tool for building container images — graduates to General Availability.
• March 2017 — Kaggle — home to the world’s largest community of data scientists and machine learning enthusiasts — is acquired and becomes part of GCP.
• March 2017 — Cloud SQL support for PostgreSQL launches in Beta. SQL Server Enterprise images on Google Compute Engine, support for Windows Server Failover Clustering (WSFC), and SQL Server AlwaysOn Availability Groups, all graduate to General Availability. Support for BigQuery to directly query Cloud Bigtable data launched.
• March 2017 — App Engine flexible environment goes into General Availability for languages Node.js, Ruby, Java 8, Python 2.7 or 3.5, and Go 1.8, along with Beta support for PHP 7.1 and .NET Core. Severless Google Cloud Functions launched into public Beta. Cloud Dataprep launched in private Beta. Regions launched in California, Montreal and the Netherlands. Number of vCPUs you can run in an instance doubled from 32 to 64 along with up to 416GB of memory.
• March 2017 — Data Loss Prevention API launched in Beta, providing access to a classification engine that includes over 40 predefined content templates for credit card numbers, social security numbers, phone numbers and other sensitive data.
• April 2017 — Kubernetes 1.6 now available to Google Container Engine users. General Availability of neural machine translation system, expanded to include Russian, Hindi, Vietnamese, Polish, Arabic, Hebrew and Thai (in addition to Chinese, French, German, Japanese, Korean, Portuguese, Spanish and Turkish.)
• April 2017 — Cloud Storage can send change notifications to a Cloud Pub/Sub topic or trigger a Cloud Function. Cloud Speech API and Spanner graduate to General Availability.
• April 2017 — MIT math professor Andrew V. Sutherland breaks the record for the largest ever Compute Engine cluster with 220,000 cores on Preemptible VMs.
• May 2017 — Northern Virginia region launched. General Availability of Compute Engine VMs with 64 virtual CPUs and up to 416 GB of memory. Cloud Source Repositories goes GA — free for up to 5 users and 50GB of storage.
• May 2017 — Cloud IoT Core, a service to securely connect your globally distributed devices to GCP and centrally manage them, launched in public beta.
• May 2017 — Google, IBM and Lyft announce Alpha release of Istio: a new open-source project that provides a uniform way to help connect, secure, manage and monitor microservices.
• June 2017 — Release of Spinnaker 1.0 (an open-source multi-cloud continuous delivery platform). Sydney and Singapore regions become available. Java 8 support added for App Engine standard.
• July 2017 — Container Engine supports Kubernetes 1.7. New London region (europe-west2) launched. Transfer Appliance launched to help ingest large amounts of data to Google Cloud Platform.
• July 2017 — GCP announces use of congestion control algorithm TCP BBR to achieve higher bandwidths and lower latencies. Web-based code editor within Cloud Shell launched in beta.

Something missing? Add a comment and I’ll update the timeline.

Further Reading

• What is Google’s Cloud Platform?
• What are the Google Cloud Platform Services?