Local and Geographic Redundancy
Within a corporation, there are two forms of redundancy:
- Local redundancy that addresses single component failures
- Geographic redundancy that addresses data center failures
Most corporations first focus on local site or data center redundancy. This ensures that if any hardware components that support mission critical applications fail, there are additional hardware components to take their place. “Take their place” means having two such hardware components configured in an Active/Active mode or having a hardware component in Hot Standby mode. Most corporations focus on local redundancy because the probability of the data center failing is less than the probability of a single hardware component failing.
Once local redundancy is achieved, geographic redundancy is sought mainly for purposes of disaster recovery. In this scenario, a second data center is acquired that contains comparable hardware. All critical applications and its data are backed up to this second data center. As with the hardware, the primary data center and secondary data center can be run in Active/Active mode (preferred) or in Active/Passive mode.
The Plexus Message Broker supports many different types of local redundancy including:
- More than one Message Server in the same instance on the same physical server
- More than one instance of the Plexus Message Broker on the same physical server
- Plexus Message Broker instances on multiple physical servers
In each of the above cases, the creation of redundant instances of the Plexus Message Broker not only provides resiliency in the service but also can provide massive throughput gain when high transaction volumes measured in hundreds of transactions per second is desired.
The Plexus Message Broker supports two types of geographic redundancy including:
- Plexus Message Brokers at the Primary and Secondary data center in Active/Passive mode
- Plexus Message Brokers at the Primary and Secondary data centers in Active/Active mode