Google Cloud, the company’s enterprise wing, has announced plans to open a second cloud region in India at Delhi by 2021. The company currently has one cloud region in Mumbai, which has been operational since 2017.
There are eight such cloud regions in Asia Pacific, and a total of 22 regions globally. So what exactly does Google mean when it says it launching a cloud region in Delhi? And how do other companies handle their cloud-based offerings? Here’s what it means.
This has nothing to do with Google’s offerings such as YouTube, Gmail or Google Maps. Their data is not hosted in these new cloud regions, but in Google’s own data centres elsewhere.
The Google Cloud regions are aimed at the company’s enterprise clients, which are large or medium companies that rely on cloud infrastructure services to run and host their applications, databases, etc. Google’s Cloud regions provide storage and services for these customers, according to the company.
Google hosts its cloud compute engine resources for clients in each of these regions. An area where this cloud infrastructure is hosted is called a region. Google explains that a ‘region’ features compute, storage, databases, security, big data, and networking services. The upcoming Google Cloud region in Delhi will also include a portfolio of key Google Cloud Platform (GCP) products.
The Delhi region will ensure that the company’s enterprise customers are able to offer lower latency to their users given the closer proximity to these critical cloud-based resources. The region will also allow geographically separate in-country disaster recovery for any mission-critical applications being hosted by Google.
Each region is further divided into three zones, though one region, which is tagged as us-central1 and located in Council Bluffs, Iowa, USA has four zones. The existing Mumbai region has three zones, while the upcoming Delhi region will also have three zones.
Interestingly, India will become second country after Japan outside US to have two regions. The maximum number of regions are hosted in the United States. Every Cloud region is connected to Google’s private infrastructure network. It also lets Google’s customers choose where they want to store their data or want to run their applications from, be it Mumbai or a region in another part of the world.
The region refers to the specific geographical location. A region typically has three zones. Inside these zones are virtual machines or zonal persistent disks. Google puts different resources in different zones. This is done to ensure isolation against any kind of physical infrastructure or software service failures.
This is a tactic followed by other cloud-based service providers as well. This is because these zones are hosting resources, which are crucial for Google’s enterprise customers, and therefore service disruption has to be minimised.
Some resources are region or zone-specific, according to Google’s own page explaining this. But other resources like images can be accessed globally, meaning they can be used or stored across any location. The physical cluster where the zones are hosted is a “distinct physical infrastructure that is housed in a data centre,” explains the page further.
“Each cluster has independent software infrastructure, power, cooling, network, and security infrastructure, and includes a large pool of compute and storage resources,” it adds. Each zone is hosted in one or more cluster.
Amazon Web Services has also divided its services into 22 regions worldwide. India has one region, which is Mumbai. Each region has several availability zones, and this is basically a collection of one or more data centres. Again a region typically has three or more zones. Amazon does not reveal the number of data centres around Mumbai nor the locations. Each zone is independently fault-tolerant and is built taking into consideration risks like floods, etc. The zone is connected with a very high speed ‘dark fibre’ connectivity given there are hosting applications for enterprise customers.