Abstract: A data processing device includes a first packet communication interface for communication with at least one host processor via a network interface controller (NIC) and a second packet communication interface for communication with a packet data network. A memory holds a flow state table containing context information with respect to multiple packet flows conveyed between the host processor and the network via the first and second interfaces packet communication interfaces. Acceleration logic, coupled between the first and second packet communication interfaces, performs computational operations on payloads of packets in the multiple packet flows using the context information in the flow state table.

Abstract: Remote transactions using transactional memory are carried out over a data network between an initiator host and a remote target. The transaction comprises a plurality of input-output (IO) operations between an initiator network interface controller and a target network interface controller. The IO operations are controlled by the initiator network interface controller and the target network interface controller to cause the first process to perform accesses to the memory location atomically.

Abstract: A method in a network element that includes multiple interfaces for connecting to a communication network includes receiving from the communication network via an ingress interface a flow including a sequence of packets, and routing the packets to a destination of the flow via a first egress interface. A permission indication for re-routing the flow is received in the ingress interface. In response to receiving the permission indication, subsequent packets of the flow are re-routed via a second egress interface that is different from the first egress interface. Further re-routing of the flow is refrained from, until receiving another permission indication.

Abstract: A network node includes one or more network adapters and a bonding driver. The one or more network adapters are configured to communicate respective data flows over a communication network by applying a transport layer protocol that saves communication state information in a state of a respective network adapter. The bonding driver is configured to exchange traffic including the data flows of an application program that is executed in the network node, to communicate the data flows of the traffic via one or more physical links of the one or more network adapters, and, in response to a physical-transport failure, to switch a given data flow to a different physical link or a different network path, transparently to the application program.

Abstract: A method for management of a host computer that includes a management controller configured to carry out, independently of the host CPU, host management instructions contained in management packets compliant with a first data link protocol. The method includes receiving the management packets from a first network operating in accordance with the first data link protocol. The management packets are encapsulated in data packets compliant with a second data link protocol, different from the first data link protocol. The data packets are transmitted to a second network, operating in accordance with the second data link protocol. The transmitted data packets are received from the second network in a network interface controller (NIC), which is installed in the host computer and connected to the second network. The NIC decapsulates the management packets from the received data packets and passes the decapsulated management packets via a sideband connection to the management controller.

Abstract: Lock access is managed in a data network having an initiator node and a remote target by issuing a lock command from a first process to the remote target via an initiator network interface controller to establish a lock on a memory location, and prior to receiving a reply to the lock command communicating a data access request to the memory location from the initiator network interface controller. Prior to receiving a reply to the data access request, an unlock command issues from the initiator network interface controller. The target network interface controller determines the lock content, and when permitted by the lock accesses the memory location. After accessing the memory location the target network interface controller executes the unlock command. When the lock prevents data access, the lock operation is retried a configurable number of times until data access is allowed or a threshold is exceeded.

Abstract: A method for network access of remote memory directly from a local instruction stream using conventional loads and stores. In cases where network IO access (a network phase) cannot overlap a compute phase, a direct network access from the instruction stream greatly decreases latency in CPU processing. The network is treated as yet another memory that can be directly read from, or written to, by the CPU. Network access can be done directly from the instruction stream using regular loads and stores. Example scenarios where synchronous network access can be beneficial are SHMEM (symmetric hierarchical memory access) usages (where the program directly reads/writes remote memory), and scenarios where part of system memory (for example DDR) can reside over a network and made accessible by demand to different CPUs.

Abstract: A method for data transfer includes receiving in an operating system of a host computer an instruction initiated by a user application running on the host processor identifying a page of virtual memory of the host computer that is to be used in receiving data in a message that is to be transmitted over a network to the host computer but has not yet been received by the host computer. In response to the instruction, the page is loaded into the memory, and upon receiving the message, the data are written to the loaded page.

Abstract: A method for data transfer includes receiving in a data transfer operation data to be written by a peripheral device to a specified virtual address in a random access memory (RAM) of a host computer. Upon receiving the data, it is detected that a page that contains the specified virtual address is marked as not present in a page table of the host computer. The peripheral device receives a notification that the page is not present and an estimate of a length of time that will be required to make the page available and selects a mode for handling of the data transfer operation depending upon the estimate.

Abstract: Peripheral apparatus for use with a host computer includes an add-on device, which includes a first network port coupled to one end of a packet communication link and add-on logic, which is configured to receive and transmit packets containing data over the packet communication link and to perform computational operations on the data. A network interface controller (NIC) includes a host bus interface, configured for connection to the host bus of the host computer and a second network port, coupled to the other end of the packet communication link. Packet processing logic in the NIC is coupled between the host bus interface and the second network port, and is configured to translate between the packets transmitted and received over the packet communication link and transactions executed on the host bus so as to provide access between the add-on device and the resources of the host computer.

Abstract: In a data network congestion control in a virtualized environment is enforced in packet flows to and from virtual machines in a host. A hypervisor and network interface hardware in the host are trusted components. Enforcement comprises estimating congestion states in the data network attributable to respective packet flows, recognizing a new packet that belongs to one of the data packet flows, and using one or more of the trusted components and to make a determination based on the congestion states that the new packet belongs to a congestion-producing packet flow. A congestion-control policy is applied by one or more of the trusted components to the new packet responsively to the determination.

Abstract: Data processing apparatus includes a host processor and a network interface controller (NIC), which is configured to couple the host processor to a packet data network. A memory holds a flow state table containing context information with respect to computational operations to be performed on multiple packet flows conveyed between the host processor and the network. Acceleration logic is coupled to perform the computational operations on payloads of packets in the multiple packet flows using the context information in the flow state table.

Abstract: A data processing device includes a first packet communication interface for communication with at least one host processor via a network interface controller (NIC) and a second packet communication interface for communication with a packet data network. A memory holds a flow state table containing context information with respect to multiple packet flows conveyed between the host processor and the network via the first and second interfaces packet communication interfaces. Acceleration logic, coupled between the first and second packet communication interfaces, performs computational operations on payloads of packets in the multiple packet flows using the context information in the flow state table.

Abstract: A method in a network element that includes multiple interfaces for connecting to a communication network includes receiving from the communication network via an ingress interface a flow including a sequence of packets, and routing the packets to a destination of the flow via a first egress interface. A permission indication for re-routing the flow is received in the ingress interface. In response to receiving the permission indication, subsequent packets of the flow are re-routed via a second egress interface that is different from the first egress interface. Further re-routing of the flow is refrained from, until receiving another permission indication.

Abstract: A method for data storage includes provisioning, in a cluster of computers, including at least first and second computers, which are connected to a packet data network, a range of RAM on the second computer for use by the first computer. Blocks of data are stored in the provisioned range for use by programs running on the first computer. Upon incurring a page fault on the first computer in response to a request for a page of virtual memory by a program running on the first computer, a block swap request is directed to the NIC of the first computer with respect to the requested page. In response to the block swap request, an RDMA read request is initiated by the NIC via the network to the NIC of the second computer, to retrieve the requested page from the provisioned range, so as to resolve the page fault.

Abstract: A method for interaction by a central processing unit (CPU) and peripheral devices in a computer includes allocating, in a memory, a work queue for controlling a first peripheral device of the computer. The CPU prepares a work request for insertion in the allocated work queue, the work request specifying an operation for execution by the first peripheral device. A second peripheral device of the computer submits an instruction to the first peripheral device to execute the work request that was prepared by the CPU and thereby to perform the operation specified by the work request.

Abstract: A method for memory access includes maintaining in a host memory, under control of a host operating system running on a central processing unit (CPU), respective address translation tables for multiple processes executed by the CPU. Upon receiving, in a peripheral device, a work item that is associated with a given process, having a respective address translation table in the host memory, and specifies a virtual memory address, the peripheral device translates the virtual memory address into a physical memory address by accessing the respective address translation table of the given process in the host memory. The work item is executed in the peripheral device by accessing data at the physical memory address in the host memory.

Abstract: A method for data transfer includes receiving in a data transfer operation data to be written by a peripheral device to a specified virtual address in a random access memory (RAM) of a host computer. Upon receiving the data, it is detected that a page that contains the specified virtual address is marked as not present in a page table of the host computer. The peripheral device receives a notification that the page is not present and an estimate of a length of time that will be required to make the page available and selects a mode for handling of the data transfer operation depending upon the estimate.

Abstract: Lock access is managed in a data network having an initiator node and a remote target by issuing a lock command from a first process to the remote target via an initiator network interface controller to establish a lock on a memory location, and prior to receiving a reply to the lock command communicating a data access request to the memory location from the initiator network interface controller. Prior to receiving a reply to the data access request, an unlock command issues from the initiator network interface controller. The target network interface controller determines the lock content, and when permitted by the lock accesses the memory location. After accessing the memory location the target network interface controller executes the unlock command. When the lock prevents data access, the lock operation is retried a configurable number of times until data access is allowed or a threshold is exceeded.

Abstract: A network node includes one or more network adapters and a bonding driver. The one or more network adapters are configured to communicate respective data flows over a communication network by applying a transport layer protocol that saves communication state information in a state of a respective network adapter. The bonding driver is configured to exchange traffic including the data flows of an application program that is executed in the network node, to communicate the data flows of the traffic via one or more physical links of the one or more network adapters, and, in response to a physical-transport failure, to switch a given data flow to a different physical link or a different network path, transparently to the application program.