Abstract: A method and apparatus are provided for managing logging data communications in a computer system. The computer system is configured with a proxy agent in communication with a client machine and storage, and an embedded agent in communication with the proxy agent and the client machine. All inbound communications from the client machine to the storage are maintained in a log local to the proxy agent. However, all outbound communications are split between two different communications paths, with one outbound path from the storage directly to the client machine and a second outbound path from the storage to the client machine via the proxy agent. One of the two outbound paths is selected based upon one or more of the characteristics of the associated communication.

Abstract: A method and apparatus are provided for managing logging data communications in a computer system. The computer system is configured with a proxy agent in communication with a client machine and storage, and an embedded agent in communication with the proxy agent and the client machine. All inbound communications from the client machine to the storage are maintained in a log local to the proxy agent. However, all outbound communications are split between two different communications paths, with one outbound path from the storage directly to the client machine and a second outbound path from the storage to the client machine via the proxy agent. One of the two outbound paths is selected based upon one or more of the characteristics of the associated communication.

Abstract: One example of this invention defines a new interface to the CIM client API that is more generic than the current standard CIM client API that clients can use for the case where they want to find all relationships between instances of A and B. The method would implement several different ways to retrieve the relationships between A and B. CIM has a concept of a qualifier, which is a tag that can be applied to supply any additional information. The provider could define in the qualifier for each association class which algorithm they think is most efficient, and the implementation of the new interface would use whichever algorithm the qualifier says will be most efficient.