Introduction and considerations
MC/ServiceGuard, or Multi-Computer/ServiceGuard, allows you to create high availability clusters of HP 9000 series computers to compose a high availability computer system that allows databases and applications to continue in spite of a hardware, software, or network failure. This highly available system detects any fault through a heartbeat mechanism and starts the recovery process automatically. We put the WebSphere master administrative repository, enterprise application database, WebSphere session database, log files, JMS providers, LDAP database, and their processes under control of this MC/ServiceGuard system to protect users from software failures as well as from failures of system processing unit, disk, or local area network (LAN) components.
As shown in Figure 12-4, we integrated, configured, and tested WebSphere end-to-end high availability system including HA firewalls that create a DMZ zone, HA Load Balancer with a hot standby LB, multiple HTTP servers with HTTP plug-in that directs requests to Web containers, multiple appservers, HA JMS server, HA LDAP server, and HA databases with a hot standby database node. Such a system eliminates single points of failure in the network, firewall, load balancer, HTTP servers, application server, administrative servers, JMS server, database server, and LDAP server.
Figure 12-4 WebSphere and MC/ServiceGuard test environment
We tested four variations for this topology:
We used Mercury LoadRunner to simulate customer site situations and to drive a heavy load for two-day duration tests, and we initiated failures in all components until we achieved satisfactory availability results. We tested administrative actions (Node Agent and Deployment Manager), servlets, JSPs, session manager, session EJBs, CMP EJBs, and BMP EJBs individually for each failure point in each component. We also tested comprehensive benchmark applications such as Trade3, Big3, and others.
We initiated temporary failures in the network, Network Dispatchers, HTTP servers, appservers (both within a node and in different nodes), administrative servers (Node Agent and Deployment Manager), and databases. We determined that the total failed client requests were a very small percentage of total client requests, which depends on frequency of failure initiation. Since we have discussed the WebSphere configuration in the previous chapters, we will focus more on HA data management and its impact on the whole WebSphere system.
WebSphere is a trademark of the IBM Corporation in the United States, other countries, or both.
IBM is a trademark of the IBM Corporation in the United States, other countries, or both.