Cloud Infrastructure
Cloud Infrastructure
After choosing the cloud service model and the cloud type offered by vendors, customers need
to plan the infrastructure architecture.
The infrastructure layer is the foundation of the cloud.
This layer consists of physical resources that are housed in Regions, Zones and Data
Centers. A Cloud provider’s IT environment is typically distributed across many regions
around the world.
A cloud region, is a geographic area or location where a cloud provider’s infrastructure is
clustered, and may have names like NA South or US East.
The cloud regions are isolated from each other so that if one region was impacted by a natural
disaster like an earthquake, the cloud operations in other regions would keep running.
Each Cloud Region can have multiple Zones (or Availability Zones or AZ for short), which are
typically distinct Data Centers with their own power, cooling and networking resources.
These Zones can have names like DAL-09 or us-east-1.
The isolation of zones improves the cloud’s overall fault tolerance, decreases latency,
and avoids creating a single shared point of failure.
The Availability Zones (and DataCenters within them) are connected to other AZs and regions,
private datacenters and the Internet using very high bandwidth network connectivity.
A cloud data center is a huge room or a warehouse containing cloud infrastructure.
These data centers contain pods and racks or standardized containers of computing resources
such as servers, as well as storage, and networking equipment - virtually everything that a
physical IT environment has.
Computing Resources: Cloud providers offer several compute options
– Virtual Servers, Bare Metal Servers, and “Serverless” computing resources.
Most of the servers in a cloud datacenter run hypervisors to create virtual servers
or virtual machines (also called VMs for short), that are software-based computers, based on
virtualization technologies.
Other servers in the racks are bare metal servers that are physical servers that aren’t
virtualized.
Customers can provision VMs and Bare Metals servers as and when they need them and run
their workloads on them.
Cloud users can also run their workloads on serverless computing resources, which are
an abstraction layer on top of virtual machines.
Storage: Information and data can consist of files, code, documents, images, videos, backups,
snapshots, and databases and can be stored in many different types of storage options on the
Cloud.
Bare Metal Servers and Virtual Servers are provisioned with default storage in local
drives.
Since these cloud servers can be provisioned and decommissioned by customers on demand
and freed up for use by other users, any information stored in a local drive can be lost when
you delete or decommission a cloud server.
However there are other storage options available on the cloud to persist data that you can
choose depending on factors like how important your data is, how quickly you want to be able
to access it, how often you access it, and how secure you need it to be.
These additional storage options include Block storage, File storage, and Object storage.
Block and file storage modes are commonly used in traditional data centers, but often
struggle with scale, performance and distributed characteristics of cloud.
Object storage is the most common mode of storage in the cloud as it’s both highly
distributed and resilient.