Emc Corporation

Walter T Dobberpuhl

Andreas L Bauer

Gregory William Lazar

Stephen J Todd

Brian R Gruttadauria

Walter A O'Brien, Iii

Ying Xie

George M Ericson

Charles H Hopkins

Timothy D Sykes

Jennifer Lyn Milliken Nicoletti

Sriram Krishnan

Munish T Desai

David Kurt Spencer

Timothy M Meggison

A method apparatus and computer program product for providing access to host attribute information in a storage area network is disclosed. The storage area network is composed of a plurality of hosts coupled to at least one initiator. Each initiator is coupled to one or more targets and each initiator has an associated identifier. In each host, the identifier of the initiator is related to other host attribute information. The identifier may be, for example, a world wide name. The host attribute information including the identifier is sent from each of the plurality of hosts to the one or more targets and stored in memory of an associated storage array. Either a host or requestor remote from the storage array may request the collected host attribute information from the storage array. A topology of the storage area network may then be formed from the host attribute information of each host.

Pushing attribute information to storage devices for network topology access

There is disclosed an improvement to a system, apparatus, method, and computer program product relating to a novel events-notification activity occurring within computer network environments including SAN, NAS and client server environments. A user at a convenient, single user-interface location in a network is enabled to establish such activity by creating and deploying template software objects which are configurable with respect to both events of interest and genre of notifications of such events. Accordingly, such user, operating at the user-interface at a client or head-end station and employing such templates, can select: (1) which particular network-failure events (from complete failures to measured degradations), such as those occurring in peripherals including disk array storage devices, are to be reported; (2) to whom such events shall be reported, worldwide; and, (3) the mode of communication by which such events shall be reported such as email, telephone, pager, etc. In those instances where certain databases in storage devices of servers employed throughout the network may contain templates which present name/content conflicts with client database templates the events-notification activity is less effective than it otherwise would be. There is disclosed a system, apparatus, method, and computer program product for handling such conflicts and thereby enhancing the effectiveness of the events-notification activity. An alternative embodiment of this solution handles the scenario where multiple networks having multiple clients sharing particular servers create template name/content conflicts.

Ambiguity-purging and template-conflict-resolution in computer network events-notification

There is disclosed a technique for achieving scalable communication within a distributed system using dynamic communication trees. In a client server environment, and particularly in such an environment in which storage systems and storage area networks are managed by distributed management software, embodiments of the present invention are presented by which communication trees are built in a manner that evenly distributes load throughout nodes in the network, whereby scalability is enhanced allowing a larger number of nodes than otherwise allowable without such trees. Other advantages achievable by utilization of such trees are also disclosed.

Scalable communication within a distributed system using dynamic communication trees

A technique for robust indication processing. In client server network environments running object oriented software, there is a need for monitoring the operating state or status of objects in the network. Prior monitoring schemes involved many specific and frequent inquiries or requests of multitudinous objects, for which the connections were constructed and destroyed on an as-needed basis, thereby taking an immense toll on the network's processing capacity. Subsequently, asynchronous monitoring has been employed as an improvement over the prior monitoring schemes, wherein an observer object monitors a subject object and is set at 'hair-trigger' to fire back indications to an interested client if and when the subject object's state changes. Asynchronous monitoring limitations include an inability to detect client and/or server failure, insecure and unreliable connections through which indications are intended to be fired back, and other drawbacks. The technique disclosed herein for asynchronous monitoring solves these problems and others by employing a heartbeat polling scheme having virtually unique ID numbers to instantaneously determine if there has been client or server failure or recycling along with a secure or persistent pipe that interconnects client and server and has the ability to tunnel through firewalls when public networks such as the Internet are utilized.

Robust indication processing

A technique for handling failure modes which may occur during the processing of indications. In a client-server environment such as a client-server network used in computer data storage running object-oriented software, the client may subscribe for certain indications on certain objects in the servers. When the state of any of those certain objects change, an indication is fired back to the client to indicate a state change. Failures in the client, server, or connection therebetween reduce the effectiveness of this state-change notification scheme. The present invention mitigates the impact of such failures, employing a technique by which a failed storage processor in a server can failover to another storage processor in the same server to enable the seamless transition of the processing of such indications to the other processor without stopping and by which the server can destroy any of its resources that were dedicated to the processing of indications for a specific client if and when such client fails.

Robust indication processing failure mode handling

There is disclosed a client-server configuration including a peripheral device wherein multiple servers interface with both the client and the peripheral. Management software for determining the operating state of the peripheral, known as server agent software, is employed in the multiple servers under control of the client. In order to avoid conflicting information reflecting multiple operating states of the peripheral, a primary/secondary algorithm is used to enable the client to determine only one peripheral operating state while allowing each server to also verify its connectivity to the peripheral and to the client.

Growing community of inventors

Milford, MA, United States of America

Walter T Dobberpuhl