The optimally redundant cluster is more reliable, multi-node deployment and requires a larger hardware investment because of the number of nodes involved. An optimally redundant cluster is the Sybase-recommended solution for larger user environments.
An optimally redundant cluster includes at least two clusters of different types (server versus data):
- An Unwired Server cluster on the corporate LAN that typically includes two Unwired Servers: a primary server and a secondary server. Each node is installed on its own host node, and would also include a Unified Agent (for remote management via Sybase Control Center)and a relay server outbound enabler (RSOE) (for relay server connections). In Sybase Control Center, you would therefore have a single entry in the Clusters folder and two entries in the Servers folder.
- An data tier cluster on the corporate LAN that typically includes two nodes: a primary node and a failover node. Typically, this is cluster is implemented with Microsoft Disk cluster or a RAID array. The data tier includes all runtime databases: a conolidated database, a messaging database, and a cluster database, each having their own purpose in supporting the runtime environment.
- At least two relay servers installed to distinct hosts on the DMZ. These relay servers can be configured as a relay server cluster if used in conjunction with an optional load balancer or other load balancing technologies.