DR Planning

Setting DC DependenciesService Restoration PrioritiesDisaster ScenariosGenerating Service Restoration Plans

Setting Data Centre Dependencies & Failovers

In the service details screen at the bottom of the service dependency tab is the Data Centre Dependencies section.

Only services that are directly dependant on the data centre service, typically infrastructure services, should have this section filled in. Other service will inherit this information from those that do.

Use the Add Data Centre / Service Failover button to add data centre dependencies for services.

For example a VM Farm may have its servers in DC1. In the event DC1 fails the VM Farm will also fail. So there is no failover.

However the VM Farm may be designed to have servers in two data centres, DC1 and DC2.

If one data centre fails the service may still work using servers in a different data centre.

In this event a failover can be specified from DC1 to DC2.

The failover can be automatic, meaning it just happens, or manual requiring configuration changes for the failover to happen.

If the service fails over from one DC to another it may impact on the performance or features of the service.

Some important IT services may have a failover designed into their architecture, in which case this would be the "preferred failover"

In the event both data centres fail other service failovers may be possible. For these use the Last Resort failover option.

  • Enter the team who would perform the failover or who can confirm the failover has happened automatically.
  • A short description of how to perform the failover.
  • Finally add the estimate number of minutes the failover process will take.

If a service is designed to fail over from DC1 to DC2, and also back again from DC2 to DC1 then enter a second line with the DC2 dependency failing back to DC1.

Setting Service Restoration Priorities

Some services are more important to the business than others.These services can be prioritised using the Restoration Priorities link in the Disaster Recovery section.

Enter the most important services that should be restored first in the event of a disaster as priority 1.

Services that a service requires to operate are automatically raised to the same priority as those requiring it.

Setting Disaster Scenarios

Various disaster scenarios can be entered using the Disaster Scenarios link in the Disaster Recovery side menu.

Enter different disaster situations affecting different data centres, either a single data centre is down or multiple.

The assumption is that data centres included in a disaster scenario are not working for whatever reason.

Enter up to 10 scenarios which may be likely based on their location or network or power dependencies.

Generating Service Restoration Plans

ServiceView can combine disaster scenarios, restoration priorities, service dependencies and data centre dependencies and failovers to generate a service restoration plan.

Each disaster scenario is modelled assuming the affected data centres are down. A list of services affected by this outage is listed first, followed by a service restoration plan describing how and when to recover each service as well as the underpinning services this service depends on.

To generate a Service Restoration Plan, firstly make sure you have completed the Restoration Priorities and Disaster Scenarios. After you have completed those select the Disaster Recovery side menu and click Complete Restoration Plan. This will generate the Restoration Plan in a new tab.