Print this Page
2.5 Availability
01. Resiliance and Availability
Close this Page

2.5.01 Resiliance and Availability
A fundamental objective of the application service is its resilience and high availability so a message as follows will never have any credibility:
"I have used the application service and when I access and go to support then I get a fatal access - I have no access to the application service".

Business Continuity
Business continuity, backup, disaster recovery and fallback are the primary business objective - nothing else is more important that continuing to provide the application service to its users.
By careful design, any single point of failure has been eliminated.   Multiple remote secure data centers are employed with multiple power supplies and more than one high speed Internet connection.   Multiple racks of resilient servers are employed with load balancing between web servers so if one web server fails, the others can continue without interruption.   Hot standby machines are ready to replace any hardware or network equipment failure.
Servers are replaced every two years to reduce the probability of failure and to reduce the carbon footprint and reduce energy costs with more effective technology.   Data is automatically replicated to many remote locations so in the event that one location is not available, business can continue from another location.

Service Resilience
The application service has an architecture with built-in resilience - in the event that one service may fail, alternative services exist to continue the business.   Functional access control is provided with nine distinct services as:
1. Public web pages that may be viewed by Google, search engines and the public who have not signed in.
2. Private services that may only be viewed by authorized users who have signed in.
3. Agent services that are only used by a user authorized as a Agent.
4. Agent-Manager services that are only used by a user authorized as a Agent-Manager.
5. Cover-holder services that are only used by a user authorized as a Cover-Holder.
6. Owner services that are only used by a user authorized as an Owner.
7. Head office services that are only used by the head office team.
8. Support services that are only used by a user who has signed in.
9. Web services that are only used by the web support team.

Service Replication
CRM services within the Agent, Agent-Manager, Cover-Holder and Owner services are replicated three different ways as:
1. CRM services using the "long page" icon on the top menu.
2. CRM services using the "hierarchy" icon on the top menu.
3. CRM services using links from the "search my data" diagram on the dashboard.

 
Resilient Architecture
This service architecture ensured that if one support service should fail, all other services will continue without impacting a users ability to do their job.   Resilience is also built into each service with three different CRM services provided to the Agent, BM, CH and Owner.   This means that if one CRM service should fail, two other methods of doing the same job continue to be provided.
For example; say the "Long Page" service has a temporary failure, the user can continue to do their work using the "Hierarchy" and the "CRM Diagram" links to access the same data.   Any user will always be able to continue and complete any job using alternative services.
This obsession with resilience and business continuity may be expensive, but it will prevent any user from reporting that they are unable to do their daily job - an alternative method or service is always available.

Continual Improvement
Continual improvements are built into the fabric of the application service to the point where changes will be made while users are actively using the application.   Users do not need to sign off while an improvement is made, people never need to reboot their computer to activate a new facility and users are always using the latest release.
At no time will the application service say "Down for Maintenance" - the application service is designed without the need for program maintenance.   Public holidays and early mornings are used to switch data centers, switch servers and to patch operating system software.   The result is an application service that continues to be provided for years and years.