There are many types of cloud services. Theses services can be hosted within organizations own infrastructure – Private Cloud with full access to all aspects. From an external an external party accessed over the internet and made available to general public – Public Cloud.
eg: Azure, Amazon Web Services.
Some organization share an infrastructure which can be thought of as Community cloud. A combination of clouds brings a Hybrid Cloud solution.
What is Cloud?
There are many definition of cloud. US NIST defines 5 critical characteristics to be a cloud i.e On demand self-service, Broad network access, Resource pooling, Rapid elasticity and Measured service.
Type of “As a Service”
When to use public cloud solutions
There is no definite right or wrong answers. Different organizations have different priorities and can operate services/datacenters at different price points. Requirements for resiliency not possible on-promise.
The key point is that Public Cloud solutions charge you based on consumption. The fact you pay only when its needed means Public Cloud fits a number of key scenarios perfectly.
- Test and development
- Disaster Recovery
- DMZ scenarios
- Special projects
- All in
How to Get Azure
- Get free on-month trail with $200 credit.
- Get Azure as part of your existing MSDN subscription
- Buy Azure as you use it
- Create an Azure agreement with Microsoft which allows creation of different subscriptions and administrators plus reduced rates
- Cloud service Provider
Account setup Methodology
Different options for creating accounts including:
- Functional teams (Sales, Legal, Marketing)
- Business Divisions (Windows, Bing)
- Geographic (North America, Europe)
Limit and Quotas
There are soft limits and hard limits. Initial subscriptions have fairly low limits to help protect you from over use.
You can increase these limits via Subscription – Usage + quotas – Request increase. On your account you can enable or remove spending limits.
Reliability Layer in the cloud.
Reliability in the hardware is what you often implement on-premise i.e Centralized storage (e.g SAN), cluster of hosts, Migration of VMs between hosts in planned situatation and failover in unplanned, typically single instance of workloads.
Reliablity in the software instance of workload i.e Distributed, multiple instances of compute and storage, VMs not migrated during maintenance.
Note: Azure datacenter architecture are highly durable and resilient. Reliability in the software is the only practical architecture for mega-scale however it does not mean its worse then reliability in the hardware.