Abstract: At any instant, a channel's path signature reflects the last successful path used to access file data. During the course of processing a request from an upstream site currently not connected to the target file, the downstream site will establish a connection to the upstream site and then include in its request response a path signature constructed by adding its signature to the channel's path signature.

Abstract: At any instant, a channel's path signature reflects the last successful path used to access file data. During the course of processing a request from an upstream site currently not connected to the target file, the downstream site will establish a connection to the upstream site and then include in its request response a path signature constructed by adding its signature to the channel's path signature.

Abstract: At any instant, a channel's path signature reflects the last successful path used to access file data. During the course of processing a request from an upstream site currently not connected to the target file, the downstream site will establish a connection to the upstream site and then include in its request response a path signature constructed by adding its signature to the channel's path signature.

Abstract: At any instant, a channel's path signature reflects the last successful path used to access file data. During the course of processing a request from an upstream site currently not connected to the target file, the downstream site will establish a connection to the upstream site and then include in its request response a path signature constructed by adding its signature to the channel's path signature.

Abstract: Large scale high performance file proxy caching sites may be configured to coalesce many client write operations into one very large assemblage of modified file data. At some point the caching site will flush all modified file data downstream towards the origin file server. In some instances the amount of modified data being flushed may be more than can be transferred in a single network request. When multiple network requests are required, the consistency guarantee provided by many filesystems requires that the file either be updated with the data contained in all of the network requests or not be modified at all. In addition, once the first flush request is processed no other file read or write requests can be serviced until the last flush request has been processed. This document discloses methods for performing atomic multi-request flush operations within a large geographically distributed filesystem environment.

Abstract: Large scale high performance file proxy caching sites may be configured to coalesce many client write operations into one very large assemblage of modified file data. At some point the caching site will flush all modified file data downstream towards the origin file server. In some instances the amount of modified data being flushed may be more than can be transferred in a single network request. When multiple network requests are required, the consistency guarantee provided by many filesystems requires that the file either be updated with the data contained in all of the network requests or not be modified at all. In addition, once the first flush request is processed no other file read or write requests can be serviced until the last flush request has been processed. This document discloses methods for performing atomic multi-request flush operations within a large geographically distributed filesystem environment.

Abstract: Objects contained within enormous geographically distributed virtual file servers spanning thousands (or even millions) of organizations are each assigned globally unique object identifiers, enabling the implementation of highly distributed indexing and retrieval operations. The file system API (application programming interface) is extended to provide a search capability. A search request targeting a specific domain creates a parallel namespace anchored in that domain's root directory. The parallel namespace, containing directories and links to all objects satisfying the search criteria, may be navigated using the standard file system API. Relevance scores, added as new members of the file attribute structure, enable the construction and presentation of views that convey where the centers of expertise associated with the search matter are located.

Abstract: The consistency callback mechanisms employed by local file systems such as NTFS and distributed file systems such as DDS, NFS and CIFS are extended to provide a shared memory foundation for efficiently broadcasting real-time high definition video from a source object to large numbers of viewers via the Internet. Distributed applications such as video viewing client applications establish connections to a common distributed file system object, and then each application registers with the underlying distributed file system to receive notifications whenever the video source modifies the source object. The data required to update images maintained by viewing clients is included in notification messages. The distributed file system employs a network of proxy cache nodes. Proxy cache nodes receive notification messages (complete with image update data) and update their cached images of the source object and then retransmit the notification messages towards the viewing clients using IP multicast techniques.

Abstract: At any instant, a channel's path signature reflects the last successful network path used to access file data. Methods whereby a distributed virtual file server, composed of standard file servers and file service proxy caches, uses path signatures to facilitate the re-routing of file system traffic to transparently overcome network link failures.

Abstract: It is not possible to maintain extremely tight synchronization of the time keeping clocks of the networked nodes comprising a distributed filesystem. However, when multiple client systems access the same file from different remote locations, a distributed consistency mechanism must ensure that all file read and write requests are only serviced from the latest version of the file. The current industry practice is to disable client-side caching when a concurrent write sharing condition arises (multiple clients active on the file and at least one of them writing). This forces all requests to flow through to the file server and consistency is maintained since all requests are then serviced from the same file image. The current practice sacrifices performance and scalability to maintain consistency.

Abstract: A method for transparently recovering from a network link failure within the context of a single network request/response operation between a client and a server is disclosed. The method requires a robust network infrastructure that provides multiple paths from the client to the server. When a network link fails, network traffic is re-routed around the failure and the request is successfully processed without the client or the server even being made aware of the failure. An error indication is returned to the client system only when the network infrastructure does not provide an alternate path that allows the request to be successfully processed.

Abstract: Large scale high performance file proxy caching sites may be configured to coalesce many client write operations into one very large assemblage of modified file data. At some point the caching site will flush all modified file data downstream towards the origin file server. In some instances the amount of modified data being flushed may be more than can be transferred in a single network request. When multiple network requests are required, the consistency guarantee provided by many filesystems requires that the file either be updated with the data contained in all of the network requests or not be modified at all. In addition, once the first flush request is processed no other file read or write requests can be serviced until the last flush request has been processed. This document discloses methods for performing atomic multi-request flush operations within a large geographically distributed filesystem environment.

Abstract: At any instant, a channel's path signature reflects the last successful path used to access file data. During the course of processing a request from an upstream site currently not connected to the target file, the downstream site will establish a connection to the upstream site and then include in its request response a path signature constructed by adding its signature to the channel's path signature.

Abstract: The consistency callback mechanisms employed by local file systems such as NTFS and distributed file systems such as DDS, NFS and CIFS are extended to provide a shared memory foundation for efficiently broadcasting real-time high definition video from a source object to large numbers of viewers via the Internet. Distributed applications such as video viewing client applications establish connections to a common distributed file system object, and then each application registers with the underlying distributed file system to receive notifications whenever the video source modifies the source object. The data required to update images maintained by viewing clients is included in notification messages. The distributed file system employs a network of proxy cache nodes. Proxy cache nodes receive notification messages (complete with image update data) and update their cached images of the source object and then retransmit the notification messages towards the viewing clients using IP multicast techniques.

Abstract: A hierarchical distributed search mechanism is integrated into a distributed file system. Traditional file system APIs (create, open, close, read, write, link, rename, delete, . . . ) and the over-the-wire protocols employed to project these APIs into remote client sites (CIFS, NFS, DDS, Appletalk) are extended to enable the dynamic creation of temporary directories containing links to objects identified by search engines (executing at sites “close” to “their” data) as meeting the search criteria specified by the first parameter of a search function call. The search function, derived from the standard file system API function create, is added to the file system API.

Abstract: A hierarchical distributed search mechanism is integrated into a distributed file system. Traditional file system APIs (create, open, close, read, write, link, rename, delete, . . . ) and the over-the-wire protocols employed to project these APIs into remote client sites (CIFS, NFS, DDS, Appletalk) are extended to enable the dynamic creation of temporary directories containing links to objects identified by search engines (executing at sites “close” to “their” data) as meeting the search criteria specified by the first parameter of a search function call. The search function, derived from the standard file system API function create, is added to the file system API.

Abstract: The consistency callback mechanisms employed by local file systems such as NTFS and distributed file systems such as DDS, NFS and CIFS are extended to provide a shared memory foundation for efficiently broadcasting real-time high definition video from a source object to large numbers of viewers via the Internet. Distributed applications such as video viewing client applications establish connections to a common distributed file system object, and then each application registers with the underlying distributed file system to receive notifications whenever the video source modifies the source object. The data required to update images maintained by viewing clients is included in notification messages. The distributed file system employs a network of proxy cache nodes. Proxy cache nodes receive notification messages (complete with image update data) and update their cached images of the source object and then retransmit the notification messages towards the viewing clients using IP multicast techniques.

Abstract: In a network of digital computers (20), a method for facilitating access by a first digital computer (24) to a file that is stored in a local file system tree (198) of a second digital computer (22). The method includes establishing, by recursively combining hierarchical file system trees (198) that are exported from one or more NDCs (50), a hierarchical domain tree (200) that encompasses digital computers (24, 26B, 26A and 22) in the network of digital computers (20). A domain manager (212) receives policy attributes, a new type of extended attributes, and enforces policies specified therein. In a particularly preferred embodiment the policy attributes identify at least one module that must be loaded by the first digital computer (24) to extend capabilities of the first digital computer (24) used in processing a file requested from the local file system tree (198) of the second digital computer (22).

Abstract: The consistency callback mechanisms employed by local file systems such as NTFS and distributed file systems such as DDS, NFS and CIFS are extended to provide a shared memory foundation for efficiently broadcasting real-time high definition video from a source object to large numbers of viewers via the Internet. Distributed applications such as video viewing client applications establish connections to a common distributed file system object, and then each application registers with the underlying distributed file system to receive notifications whenever the video source modifies the source object. The data required to update images maintained by viewing clients is included in notification messages. The distributed file system employs a network of proxy cache nodes. Proxy cache nodes receive notification messages (complete with image update data) and update their cached images of the source object and then retransmit the notification messages towards the viewing clients using IP multicast techniques.

Abstract: Systems and methods for extending the functionality of the consistency callback mechanisms employed by local file systems such as NTFS and distributed file systems such as DDS, NFS and CIFS so as to provide a shared memory foundation for the support of highly interactive distributed applications are disclosed. Distributed applications establish connections to a common distributed file system object; each application registers the file system events for which it wants to receive a notification (either synchronously or asynchronously). In this manner, each application becomes almost instantly aware of the file access activities for which it registers and which are being performed by the other applications.