Configuring HP DCE Cells

Integrating DCE Services with MC/ServiceGuard

While the replication mechanisms of the Security Service and the Naming Service differ in design and implementation, they share this master-slave approach. Therefore, while both services can be considered highly available for read operations, they do present a single point of failure for write operations.

The Time Service, on the other hand, does not present the same level of vulnerability. Most mission critical installations will configure more than the minimum necessary time servers with multiple time providers. This being the case, the loss of a single time server is usually not critical.

Installations should not establish an MC/ServiceGuard configuration for the purpose of maintaining the Time Service alone.

DCE-based applications can also present a single point of failure, unless the developers provide for the replication of data and functions between multiple servers. Since replication is a complicated and complex process, many application designers may choose to depend on a “fail over” approach such as MC/ServiceGuard to provide availability, rather than develop and maintain their own replication mechanisms.

In summary, you only need to use MC/ServiceGuard to increase the availability of DCE Core Services and DCE-based services that are not replicated. In the case of the DCE Core services, these are the write functions provided by the DCE master Security and Naming replicas.

Planning Considerations

In your planning for MC/ServiceGuard, you must consider the following characteristics of DCE and DCE-based programs:

•The DCE runtime and daemons do not themselves support the concept of dynamic IP addresses.

•Normal DCE programming practice assumes that all IP addresses on the host should be used for endpoints for exported services.

The DCE runtime determines the available IP addresses on the node during the execution of any of the rpc_server_use_* routines. These routines are used in every DCE server to select the protocols over which the server will provide services. A side effect of this call is that the list of IP addresses supported by the node is established for use later when determining the binding vector. When this vector is obtained by a server main routine and registered in the endpoint map, the endpoint map will contain entries for every IP address identified earlier during the rpc_server_use_* call. In addition, should this binding vector be

Planning and Configuring HP DCE 1.7

5-31