Active Cluster Guide > Introducing Active Cluster > Active Cluster Architecture
 
Active Cluster Architecture
An Active Cluster is initially set up on a computer running TDV. Additional computers (nodes) running TDV can then join the Active Cluster. Each cluster node is identical to a standalone TDV server in terms of system configuration and repository support.
All cluster nodes must be running the same version and patch level of Active Cluster. See the TDV Installation and Upgrade Guide for more information.
Active Cluster uses a share-nothing architecture, so you can mix UNIX and Windows servers in a cluster. However, we recommend that all nodes be of comparable operational characteristics: memory, number of CPUs, and CPU speed.
All servers are peers. Caching, statistics, and triggers are executed in a round-robin sequence.
Active Cluster characteristics:
Metadata Repository
Cache Data Sharing
Using a Load Balancer
About the Timekeeper
Triggers
Propagation of Metadata Changes
Error Recovery
Joining and Leaving an Active Cluster
Security Implementation