Google BigQuery is the Mountain View, CA search giant's data warehouse offering. But it's not your typical data warehouse (DW). Most of BigQuery's competitors are essentially cloud-hosted Massively Parallel Processing (MPP) data warehouse products. BigQuery, on the other hand, was born of an internal project at Google called Dremel, that was essentially a log analytics platform.
Also read: Google's BigQuery goes public
Also read: Looker and Google team up around BigQuery Machine Learning
BigQuery is serverless, so customers needn't worry about sizing or provisioning discrete clusters. It also leverages advanced networking technology that maximizes bandwidth between compute nodes. While BigQuery can handle DW workloads, and while Google absolutely promotes its use in such scenarios, it's designed for Web scale, and not just Enterprise scale, analytics work.
BigQuery is popular, but not everyone can use it. With "data sovereignty" regulations (wherein the government corresponding to a customer's nationality requires the customer's data to remain physically on servers in that country) on the rise, many cloud services can only be used if they're hosted, and use storage resources, in the customer's country or region.
With that in mind, at its Cloud Next event in London today, Google is announcement of BigQuery in that very region. The UK joins the US and Japan (Tokyo, specifically) in the family of countries where BigQuery is hosted. Google's Jordan Tigani, Engineering Director on the BigQuery team, also explained to me that other geographies, including Germany and Finland, will be coming online in the coming months. And BigQuery will be coming online in Sydney, Australia, this coming Monday, October 15th.
This geospatial is Web scale
Speaking of geography, Google had announced the alpha release of new geospatial features for BigQuery, back at its recent Cloud Next US event in July at Moscone Center in San Francisco. The company is today promoting that feature set from alpha to beta.
Officially dubbed BigQuery Geographic Information Systems (GIS), the feature set adds a GEOGRAPHY data type, new geography functions in standard SQL (BigQuery's SQL dialect) and a visualization tool called BigQuery GeoViz.
Doing GIS operations at scale can be a tricky business, but scale is what Google sees as BigQuery's sweet spot. Google will need to push on unique features like GIS and its serverless architecture, in order to compete with the likes of cloud DW competitors Redshift, Azure SQL Data Warehouse and Snowflake. It will also need to deploy the service across existing and future Google Cloud Platform (GCP) regions. Clearly, the company is taking both challenges seriously, with today's announcements.