摘要:
A method for directory enabled, self service, single sign on user account management, the method implemented in conjunction with a directory, the directory comprising resource directory entries and user directory entries. Exemplary embodiments include receiving, from a user through a browser, an SSO user ID and SSO password of the user, the SSO password being known only to the user, retrieving from the directory, in dependence upon the SSO user ID and SSO password, a user directory entry for the user, and creating, in dependence upon the user directory entry, a new mapping of a resource to the SSO user ID.
摘要:
Access to target data processing systems frequently requires a password to be submitted in conjunction with user identification. The required rules and syntax for such passwords may vary widely from system-to-system and the number and variety of systems makes password management difficult. An analysis of an initially assigned or known valid password is performed, and the nature of each character within the password is defined, i.e., a numeric character, a punctuation character, a lower-case alphabetic character, or an upper-case alphabetic character. Randomly generated characters of identical nature are then assigned to each position within the password to create a new password, which will comply with the specified rules and syntax for a particular target data processing system.
摘要:
A clustered computer system provides both speed and reliability advantages. However, when communications between the clustered computers is compromised those same computers can become confused and corrupt database files. The present method and apparatus are used to improve the management of clustered computer systems. Specifically, the system expands the number of nodes available for failover conditions.
摘要:
One application of clustered computer systems is to support failover of applications and shared resources. Another is to support scalable or fault-tolerant distributed applications. The present invention utilizes a higher-level clustering mechanism (a multi-cluster) overlayed on top of multiple underlying clusters (subclusters) to extend their capabilities. In the described embodiment, subclusters supporting application and shared resource failover across a smaller number of nodes is overlayed with a multi-cluster supporting a larger number of nodes. The multi-cluster manages cluster-level communication among nodes, and the subclusters directly manage only the shared device and application resources which they are permitted to control. It is possible to move resources between nodes which reside in different subclusters. The multi-cluster layer also externalizes interfaces similar to those of the subclusters, providing application compatibility.