High Availability and Disaster Recovery

Load Balancing

• A load balancer is designed to distribute load between multiple devices

• Provides fault tolerance.

• A Virtual IP address is an IP address that does not change as compared to the IP address on the physical network interface card. 

• Active/active - Distribute the load equally between all active servers.

• Active/passive - Distribute load to the active system only and fail-over to the standby system if the active system is offline.

Multipathing

• Ensuring there are multiple paths between a source and destination across a network.

• If one path is not available, the traffic can be redirected using an alternative/backup path to reach the destination.

Network interface card (NIC) teaming

• NIC Teaming allows you to group multiple NICs into a single software-based virtual network adapter.

• NIC Teaming allows a server with multiple NICs to function in active/standby.

• This allows both NICs to function by a logical network interface card to provide high bandwidth.

Redundant hardware/clusters

• Fault tolerance is ensuring if a failure happens, a system is will able to operate.

• To achieve fault tolerance, redundancy of systems and components can be used to ensure a system or network to remain operational.

• Multiple storage devices, network interface cards, power supply units, network devices, internet connections, etc.

• Switches

• Routers

• Firewalls

Facilities and infrastructure support

• Uninterruptible power supply (UPS) - Provides additional power for a short time to a device after a power lose occurs.

• Power distribution units (PDUs)

• Generator

• HVAC - Heating, Ventilating and Air Conditioning (HVAC), HVAC systems are usually implemented within a Data Center to ensure devices are kept cool at all times. HVAC system can be integrated into fire suppression system.

• Fire suppression

Redundancy and high availability (HA) concepts

Hot site

• This is a duplication of the organizations such as hardware, software and data.

• This type of site has everything duplicated and ready-to-go when needed.

Warm site

• This types of recovery site may have server racks and is just enough to be up and running.

• The company has to bring their data and hardware resources.

Cold site

• This site has no hardware at all.

• It’s an empty building or room space.

• The company will be required to bring their data, software, hardware, network components and their employees.

Cloud site

• The cloud site allows an organization to use a cloud provider's infrastructure to sustain their business during a disaster.

• Provides scalability and redundancy to the customer.

• Uses a pay-as-you-go model for the resources used by the customer.

Active-active vs. active-passive

• Multiple Internet service providers (ISPs)/diverse paths ensure an organization uses multiple ISPs to ensure internet access is always available when needed.

First Hop Redundancy Protocol (FHRP)/Virtual Router Redundancy Protocol (VRRP)

• FHRP allows two (2) or more routers share a virtual IP address and a virtual MAC address which will allow those routers to act as a single virtual router on a network as the default gateway.

• Virtual Router Redundancy Protocol version 2 (VRRPv2) - A non-proprietary election protocol that dynamically assigns responsibility for one or more virtual routers to the VRRP routers on an IPv4 LAN. This allows several routers on a multi-access link to use the same virtual IPv4 address.

• VRRPv3 - Provides the capability to support IPv4 and IPv6 addresses. VRRPv3 works in multi-vendor environments and is more scalable than VRRPv2.

Mean Time Between Failures (MTBF)

• This is a predicted time between a system failures. 

• This provides an idea of how to better predict and/or plan for a possible failure of a system.

Mean Time To Restore (MTTR)

• The mean-time-to-restore (MTTR) is the mean time is takes to repair a system to a working state. 

Mean Time to Failure (MTTF)

• This is the mean expected lifespan of a system that is non-repairable.

• If a failure occur, the system must be replaced. 

Recovery Time Objective (RTO) 

• The Recovery Time Objective (RTO) defined how quickly can you get the system up and working.

Recovery Point Objective (RPO)

• While recovering a system, some services and data may not be available while the system may be up and running. 

• The Recovery Point Objective (RPO) is used to define at what level a recovered system is considered to be acceptable. 

Network device backup/restore

• When considering to backup and restoration, it important to have a clear understanding of the state of the device and it's configurations.

• The actual state is the current running configurations on the device.

• The perceived state is simply the perception of the system or device.

• The desired state is simply the preferred state of the system or device.

• When restoring data/configurations to a device, the restoring may not archive the desired state.