In business, outages and downtime cost you and your customers valuable time and money.
Making the decision to move your resources from on-premises (your control) to the cloud (someone else’s control) requires thought, planning and confidence in your cloud partner. IBM Cloud provides infrastructures designed with built-in high availability features and additional features to customize your solution to your needs.
In our second blog post, we introduced the infrastructure architecture and the basic Terraform scripts for creating an infrastructure. Now we are going to talk about the planning that needs to be done before you create an infrastructure. Your preparation and deployment depend on your answers to the five W’s: Why, What, When, Where, Who and How.
Why do you need a resilient infrastructure?
Your resources are your livelihood. You need your data and applications to be available at all times, with limited outages. Outages come in many forms:
- Power loss
- Planned or unplanned maintenance
- Hardware failure
- Network failure
- Natural disaster
You can protect against power outages and network failures by ensuring you do not have a single point of failure. Redundant resources help protect against hardware failures and planned or unplanned maintenance and, to some extent, natural disasters. Robust backup and restore plans help protect against ransomware attacks.
IBM Cloud infrastructures are designed to avoid single point of failure by providing the following:
- Redundant power feeds
- Redundant network devices and connections
- Redundant systems
These redundancies are automatically provided in the IBM Cloud underlying infrastructure.
What types of resources do you have?
The infrastructure you build on the cloud depends on the kinds of data and application access that you need. For your resources, consider these questions:
- If there is a disruption, how quickly do you need to recover to meet any SLAs?
- Is the data or application for archival purposes and can wait?
- Is the data or application business critical, and do you need immediate access to a standby system/database?
- Are your workloads consistent over time or do they fluctuate with holidays or special events?
- Are your load demands constant or do you need dynamic adjustments at key times?
IBM Cloud offers infrastructures for all use cases, from routine data storage to business-critical data and applications. There are also features to add that automatically scale the infrastructure based on load.
When do you use the resources?
The frequency with which you access your data and applications is important to consider when you are planning your infrastructure. How often do you use this data/application? Can you go minutes, hours or days without access, if necessary? Is this data or application used for any of the following?
For archival or once-a-year reporting: Is a single point of failure tolerable? Is a single virtual instance sufficient?
Semi-regular functions: Is a redundant setup needed? Can the data and application be in a single availability zone?
For core functions: Are multiple zones in a region sufficient redundancy or do you need resources across multiple zones?
For critical processes: Is the data or application accessed daily or hourly? Can any downtime be tolerated? Do you need an active/standby configuration or an active/active configuration?
IBM Cloud offers a variety of availability zone options to fit your specific resources needs.
Where do you want to deploy your resources?
How you want your resources deployed depends on what types of resources you have:
Single-availability zone: One zone, no redundancy — good for data/applications that are rarely used or accessed. Resiliency is not the main concern. A well-implemented backup and restore process protects the data.
Multi-zone region: Multiple availability zones in one region — good for most business applications and data. Redundant virtual server instances in multiple places and multiple connections keep data and applications accessible.
Cross multi-zone region: Multiple zones across different regions — good for critical systems and data and protects against localized weather situations and power failures. With active/standby and active/active configurations, any disruptions are negligible.
IBM Cloud offers infrastructures that have basic resiliency designed into them and multiple interfaces for manual and automated deployment.
Who provides features to enhance your resiliency?
In addition to built-in resiliency, IBM Cloud provides additional features that you can use to enhance your resiliency:
Auto scale: Using auto scale, you can set a minimum and maximum number of virtual server instances to have based on the system load. Virtual server instances are dynamically created and removed based on load.
Anti-affinity: Using anti-affinity, you can specify how your virtual server instances are spread out in the data center, either by hypervisor or power spread. Virtual server instances are distributed based on hypervisor location or power source.
IBM Cloud provides manual and automated interfaces for you to build your infrastructures and add or subtract features that you want.
How does IBM Cloud help you create your infrastructure quickly and accurately?
IBM Cloud builds in a basic level of resiliency and high availability by providing the following:
- Multiple power feeds
- Fiber links
- Dedicated generators
- Battery backup
IBM Cloud also offers multiple ways to configure your infrastructure:
IBM Cloud user interface: Manually configure your VPC and virtual server instances and other components with the IBM Cloud interface.
Automation: Create infrastructures accurately and quickly, with limited manual errors:
Terraform scripts: Running scripts on your local machine, you can create simple and complex infrastructures for single availability zones, multi-zone regions and cross multi-zone regions. You are able to edit variable files and run scripts that create the VPC and components. Best for people who are comfortable using and editing code files.
IBM Cloud Schematics interface: With IBM Cloud Schematics, you provide code URLs and parameter values and the system executes the underlying Terraform scripts. Some familiarity with code is needed.
IBM Cloud catalog tile: You enter parameter values and the system executes the underlying Terraform scripts. No coding experience is needed.
To get started, look at your resources to see what types you have and how often you access them. Determine your infrastructure size and location: single zone, multi-zone or cross-multi-zone. For business-critical resources, consider using active/active or active/standby configurations.
Get started the easy way using the High Availability tiles for MZR and SAZ in the IBM Cloud Catalog. From the tiles you can do the following:
- Create an infrastructure through the user interface
- Deploy your infrastructure with IBM Cloud Schematics
- Download the Terraform scripts to run on your local machine