Riverbed Technology, Inc.

Fastforward Networks, Inc.

Nbt Technology, Inc.

Yahoo! Inc.

Google Inc.

Conviva Inc.

Internap Network Services Corporation

Other

Steven McCanne

Michael Demmer

David Tze-Si Wu

Alfred Landrum

Lap Nathan Trac

Andrew Swan

Loris Degioanni

Nitin Gupta

William J Welch

Arvind Jain

Christopher James White

Hui Joyce Zhang

Aditya Ravikumar Ganjam

Ion Stoica

Jibin Zhan

MArk Stuart Day

Kand Ly

Dimitri Stratton Vlachos

Kartik Subbanna

Daniel Edwin Burman

John Martin

James E Mace

Daniel McKernan

Jeremy Johnson

Eric Klinker

Nitin Parab

James Eric Klinker

Henri Dubois-Ferriere

William DeStein

Charles Fraleigh

Nicholas Thompson

Abhay K Parekh

David B Anderson

Dmitri S Vlachos

William C Destein

Derek J Watson

Gordon Chaffee

Matthew G Podolsky

Kevin J Lang

Michael J Demmer

Virtual storage arrays consolidate branch data storage at data centers connected via wide area networks. Virtual storage arrays appear to storage clients as local data storage; however, virtual storage arrays actually store data at the data center. The virtual storage arrays overcomes bandwidth and latency limitations of the wide area network by predicting and prefetching storage blocks, which are then cached at the branch location. Virtual storage arrays leverage an understanding of the semantics and structure of high-level data structures associated with storage blocks to predict which storage blocks are likely to be requested by a storage client in the near future. Virtual storage arrays determine the association between requested storage blocks and corresponding high-level data structure entities to predict additional high-level data structure entities that are likely to be accessed. From this, the virtual storage array identifies the additional storage blocks for prefetching.

Virtualized data storage system architecture

In a network including WAN accelerators and segment-oriented file servers, a method comprises responding to a client request to manipulate a file via a network file protocol by receiving a first request at a first WAN accelerator, wherein the request is a request to open a file located at a file server that is a segment-oriented file server, sending a local request for the file, corresponding to the first request, from the WAN accelerator to the file server, using a segment-aware network request protocol, returning at least a portion of the requested file in the form of a representation of a data map corresponding to the at least a portion of the requested file stored on the file server and using a data map for reconstruction of the requested file.

Hybrid segment-oriented file server and WAN accelerator

Network traffic information from multiple sources, at multiple time scales, and at multiple levels of detail are integrated so that users may more easily identify relevant network information. The network monitoring system stores and manipulates low-level and higher-level network traffic data separately to enable efficient data collection and storage. Packet traffic data is collected, stored, and analyzed at multiple locations. The network monitoring locations communicate summary and aggregate data to central modules, which combine this data to provide an end-to-end description of network traffic at coarser time scales. The network monitoring system enables users to zoom in on high-level, coarse time scale network performance data to one or more lower levels of network performance data at finer time scales. When high-level network performance data of interest is selected, corresponding low-level network performance data is retrieved from the appropriate distributed network monitoring locations to provide additional detailed information.

Distributed network traffic data collection and storage

Systems and techniques are described for optimizing communications between a client and a server. Specifically, in some embodiments, an executing script on a client can send a resource request to a server. In response, the server can send an optimized version of the resource back to the client. The client can then reconstruct the resource from the optimized version of the resource.

Optimizing communications using client-side reconstruction scripting

Delivering information over a distribution network using a distribution manager is disclosed. A first assignment of one or more sources peers in the distribution network is sent to a destination peer in the distribution network. At least one of a change in a network condition and a policy change is detected. It is determined that at least one modification should be made to the first assignment. A second assignment of one or more source peers in the distribution network is sent to the destination peer.

Reassigning source peers

In address-manipulation enabled transaction accelerators, the transaction accelerators include outer-connection addressing information in packets emitted over an inner connection between transaction accelerators and inner-connection addressing information is added in packets sent over the inner connection. The inner-connection addressing information can be carried in TCP option fields, directly in other fields, or indirectly through data structures maintained by the endpoints processing the connection. Address information can be encoded into header fields originally intended for other purposes but that are unused or encoded into used fields, overlaid in combination with other data that is being carried in those used fields. The existence of inner-connection addressing information in a packet can be signaled by a flag in the packet, by a bit or other designated encoding. The flag can be in an unused header field or overlaid. Where replacement and option addition is needed, swappers and unswappers might be used.

Address manipulation to provide for the use of network tools even when transaction acceleration is in use over a network

An overlay protocol and system for allowing multicast routing in the Internet to be performed at the application level. The overlay protocol uses 'native' Internet multicast and multicast routing protocols to route information, according to overlay routing tables. Overlay groups are mapped to native multicast groups to exploit native multicasting in regional or local forwarding domains. Use of the overlay protocol allows overlay distribution to be handled in a more intelligent and bandwidth-managed fashion. Overlay routers are placed at each of several local area networks, Internet service provider's point of presence, enterprise, or other cohesively-managed locations. The overlay computers are configured according to bandwidth and security policies, and perform application-level multicast distribution across the otherwise disjoint multicast networks by using the overlay routing. The result is an overlay multicast network that is effectively managed according to local network management policies. Application-level control can be applied to the transferred data at the overlay routers.

Performing multicast communication in computer networks by using overlay routing

A method and apparatus are provided for scheduling a heterogeneous communication flow. A heterogeneous flow is a flow comprising packets with varying classes or levels of service, which may correspond to different priorities, qualities of service or other service characteristics. When a packet is ready for scheduling, it is queued in order in a flow queue that corresponds to the communication flow. The flow queue then migrates among class queues that correspond to the class or level of service of the packet at the head of the flow queue. Thus, after the head packet is scheduled, the flow queue may be dequeued from its current class queue and requeued at the tail of another class queue. If the subsequent packet has the same classification, it may be requeued at the tail of the class queue or may remain in place for another servicing round.

Method and apparatus for scheduling a heterogeneous communication flow

In a network supporting transactions between clients and servers and proxies that are interposable in a network path, a pair of proxies can modify a packet stream such that packet data from client to server is transformed at a client-side proxy of the pair and untransformed at a server-side of the pair and packet data from server to client is transformed at the server-side proxy and untransformed at the client-side proxy. A discovering proxy transparently discovers its position in a proxy pair by using proxy signals. A discovering proxy might determine that it is a client-side proxy by receipt of a packet from client without a proxy signal, or that it is a server-side proxy by receipt of a packet from server without a return proxy signal. Once a proxy pair is discovered, it might transform traffic from server to client or vice versa, transforming and untransforming the traffic.

Cooperative proxy auto-discovery and connection interception

A system of network proxies distributes data to multiple servers. Each network proxy is associated with a server. A network proxy intercepts a client request for data. If the network proxy determines that the request can be served using a copy of data stored on the local server, rather than the data stored on a remote server, it diverts the request to the local server. If the network proxy determines that the request cannot be served using a data from the local server, the network proxy diverts the request to a remote server storing the primary copy of the data. A server map specifies the locations of the primary copies of data. When a primary copy of data is updated on one of the servers, the associated network proxy propagates the updated data to the other servers. The servers can provide data from files, e-mail services, databases, or multimedia services.

Reliability and availability of distributed servers

Identifying related network traffic data for monitoring and analysis

Multilevel network monitoring system architecture

Growing community of inventors

Berkeley, CA, United States of America

Steven McCanne