Sun Microsystems, Inc.

Vmware, Inc.

Oracle America, Inc.

Jochen Behrens

Marcelino M Dignum

Wayne F Seltzer

William T Zaumen

Yong Wang

Xinhua Hong

John P Petry

Biswadeep Nag

Yu Ying

Santiago Martin Pericas-Geertsen

Ariel Hendel

Rabin Sugumar

Dexiang Wang

Michael K Wong

Awan Kumar Sharma

Stephen Phillips

Sourabh Bhattacharya

Deepika Kunal Solanki

Hugh Kurth

Sarthak Ray

Pankaj Gupta

Ajoy Siddabatuni

Suraj Sudhir

Some embodiments provide a method for a data message processing device that includes multiple network interfaces associated with at least two different non-uniform memory access (NUMA) nodes. The method receives a data message at a first network interface associated with a particular one of the NUMA nodes. Based on processing of the data message, the method identifies multiple output options for the data message. Each of the output options has an equal forwarding cost and each output option is associated with a respective one of the NUMA nodes. The method selects an output option for the data message that is associated with the particular NUMA node to avoid cross-NUMA node processing of the data message.

NUMA-aware output interface selection

Example methods and systems for multicast packet handling based on flow cache information are described. In one example, a network element may configure flow cache information associated with a multicast flow. The flow cache information may specify a set of actions that is configured based on a sequence of function calls. In response to detecting a multicast packet associated with the multicast flow, fast-path processing may be performed based on the flow cache information. This may include executing a replication action to generate a first packet replica and a second packet replica. First processing action(s) may be executed to process the first packet replica to generate and send a first output packet towards a first multicast destination. Second processing action(s) may be executed to process the second packet replica to generate and send a first output packet towards a second multicast destination.

Multicast packet handling based on flow cache information

Described herein are systems, methods, and software to manage replay windows in multipath connections between gateways. In one implementation, a first gateway may receive a packet directed toward a second gateway and identify a path from a plurality of paths to the second gateway. Once identified, the first gateway may increment a sequence number associated with the path and encapsulate the packet with a unique identifier for the path in the header with the incremented sequence number. The first gateway the communicates the encapsulated packet to the second gateway.

Managing replay windows in multipath connections between gateways

Example methods and computer systems for receive-side processing for encapsulated encrypted packets. One example may comprise: in response to receiving, over a tunnel, a first encapsulated encrypted packet that includes a first encrypted inner packet and a first outer header, generating a first decrypted inner packet by performing decryption and decapsulation; and based on content of the first decrypted inner packet, assigning the first decrypted inner packet to a first processing unit. The method may further comprise: in response to receiving, over the tunnel, a second encapsulated encrypted packet that includes a second encrypted inner packet and a second outer header, generating a second decrypted inner packet by performing decryption and decapsulation; and based on content of the second decrypted inner packet, assigning the second decrypted inner packet to a second processing unit, thereby distributing post-cryptography processing over multiple processing units.

Receive-side processing for encapsulated encrypted packets

A server interconnect system includes a first server node operable to send and receive messages and a second server nodes operable to send and receive messages. The system further comprises a first interface unit in communication with the first server node and a second interface unit in communication with the second server node. The first interface unit has a first set of message send registers and a first set of message receive registers. Similarly, the second interface unit has a second set of message send registers and a second set of message receive registers. The server interconnect system also includes a communication switch that receives and routes a message from the first or second server nodes when either of the first or second registers indicates that a valid message is ready to be sent. A method implemented by the server interconnect system is also provided.

Simple, reliable, connectionless communication mechanism

A method and apparatus for accelerating processing of a structured document. A hardware XML accelerator includes one or more processors (e.g., CMT processors), one or more hardware XML parser units, one or more cryptographic units and various interfaces (e.g., to memory, a network, a communication bus). An XML document may be processed in its entirety or may be parsed in segments (e.g., as it is received). A parser unit parses a document or segment character by character, validates characters, assembles tokens from the document, extracts data, generates token headers (to describe tokens and data) and forwards the token headers and data for consumption by an application. A cryptographic unit may enforce web security, XML security or some other security scheme, by providing encryption/decryption functionality, computing digital signatures, etc. Software processing, bus utilization and latencies (e.g., memory, bus) are greatly reduced, thereby providing significantly improved XML processing and security processing throughput.

Method and apparatus for hardware XML acceleration

A hardware unit for parsing an XML document includes embedded logic or circuitry for accessing the document, decoding it to change a character set, validating individual characters of the document, extracting tokens, maintaining a symbol table and generating binary token headers to describe the document's structure and convey the document's data to an application. Tokenization, the process of identifying tokens and generating token headers, may be controlled by a finite state machine that recognizes XML delimiters in the document's markup and activates state transitions based on the current state and the recognized delimiter. The parser unit may be implemented within a hardware XML accelerator that includes a processor, a DMA engine, a cryptographic engine, memory (e.g., for storing a document, maintaining a symbol table) and various interfaces (e.g., network, memory, bus).

Hardware unit for parsing an XML document

A method and apparatus for performing virtualized parsing of an XML document. A document is divided into multiple segments, which may correspond to separate packets containing portions of the document, disk blocks, memory pages, etc. For each segment, a processor operating within an XML accelerator initiates parsing by identifying to a hardware parsing unit the document segment, a symbol table for the document and a location for storing state information regarding the parsing. Each segment is parsed in sequence, and the state information of the parsing is stored after each segment is completed, for retrieval when the next segment is to be parsed.

Method and apparatus for virtualized XML parsing

A hardware finite state machine for facilitating the processing of an XML (Extensible Markup Language) document or other structured data stream. An accelerator is implemented in hardware to enable fast processing of a document (or a segment thereof). The accelerator includes a finite state machine that embodies a ternary CAM (Content-Addressable Memory) and associated RAM (Random Access Memory). Processing of the document is divided into multiple states, with each state transition defined by a markup delimiter that triggers the transition. The CAM is programmed with entries containing the processing states and, for each possible transition from that state, a pattern for matching delimiters that trigger the possible transitions. For a CAM entry matching the current processing state and a sequence of characters from the document, which may contain a delimiter, the associated RAM identifies the next state and any action to be taken (e.g., to shift the sequence of characters).

Programmable hardware finite state machine for facilitating tokenization of an XML document

An apparatus and method for managing the receipt of communication traffic in the form of packets or other units. The apparatus includes a communication interface (e.g., a NIC, a TCA) coupled to one or more host computer systems. Through Direct Memory Access (DMA) operations, the interface reassembles payloads of received packets into host buffers based on their sequence numbers, without buffering them in the interface. Packet headers are separated from the payloads and passed to a host for protocol processing after the payload DMA is completed. Host buffers may be of virtually any size. For each communication connection, state information is maintained on the interface, which may identify an upper level protocol so that an upper level protocol header is passed to the host as part of the packet header, not as part of the payload. Protocol termination remains in the host.

Growing community of inventors

Santa Cruz, CA, United States of America

Jochen Behrens