Mellanox Technologies, Ltd.

Amir Ancel

Shay Aisman

Natali Shechtman

Yuval Itkin

Amit Mandelbaum

Ido Gilboa

Sagi Arieli

Oren Vaserberger

Doron Haritan Kazakov

Iftah Levi

Ran Sofer

Liran Liss

Idan Burstein

Itamar Rabenstein

Oren Duer

Eliav Bar-Ilan

Ran Avraham Koren

Achiad Shochat

Charlie Mbariky

Guy Twig

Yossi Kendel

Tal Gilboa

Gil Rockah

Ido Gross

A network device (ND) includes packet processing circuitry and performance optimization circuitry. The packet processing circuitry is connected to a network and is configured to process communication packets for communicating over the network. The packet processing circuitry includes a plurality of configuration registers for setting one or more operation parameters of the ND. The performance optimization circuitry is configured to improve a performance measure of the ND by iteratively calculating the performance measure and adjusting values of one or more of the configuration registers based on the performance measure.

Real time performance tuning of network devices

An apparatus includes operational circuitry and Hardware Diagnostics Circuitry (HDC). The HDC is configured to receive a definition of multiple trigger rules, each trigger rule specifying a respective trigger event as a function of trigger data sources in the operational circuitry, to receive a definition of (i) a pre-trigger logging set selected from among a plurality of diagnostics data sources in the operational circuitry, and (ii) for each trigger rule, a respective post-trigger logging set including a set of one or more of the diagnostics data sources, and, during operation of the operational circuitry, to log the diagnostics data sources in the pre-trigger logging set, to log the trigger data sources and to repeatedly evaluate the trigger rules, and, in response to triggering of a given trigger event by a given trigger rule, to start logging the diagnostics data sources in the post-trigger logging set of the given trigger rule.

High-frequency event-based hardware diagnostics

A method includes receiving a plurality of rules for processing packets arriving at at least one packet-processing device, the rules being based on multiple parameters associated with the packets. The method further includes causing the packet-processing device to process multiple subsets of the packets using different respective sequences of packet-processing instructions, which implement the rules via different respective parameter-value-ascertaining orders for ascertaining respective values of the parameters for each of the packets. The method further includes, based on the processing of the subsets, computing respective scores for the sequences of packet-processing instructions. The method further includes, based on the scores, selecting one of the sequences of packet-processing instructions. The method further includes causing the packet-processing device to process the packets using the selected sequence of packet-processing instructions. Other embodiments are also described.

Dynamic processing trees

In one embodiment, data communication apparatus includes a network interface including one or more ports for connection to a packet data network and configured to receive content transfer requests from at least one remote device over the network, a storage sub-system to be connected to local peripheral storage devices, and including at least one peripheral interface, and a memory sub-system including a cache and RAM, and processing circuitry to manage transfer of content between the remote device(s) and the local peripheral storage devices via the peripheral interface(s) and the cache, responsively to the content transfer requests, while pacing commencement of serving of respective ones of the content transfer requests responsively to a metric of the storage sub-system so that while ones of the content transfer requests are being served, other ones of the content transfer requests pending serving are queued in at least one pending queue.

System for improving input / output performance

Apparatus for communications includes a CPU, a system memory, and a network interface controller (NIC), which is configured to receive incoming data packets from a network, to post the received data packets in a designated queue for delivery to the CPU. The NIC issues interrupts to the CPU in response to the incoming data packets at a rate determined, for the designated queue, in accordance with an interrupt moderation parameter that is set for the queue. During each of a succession of monitoring periods, the CPU measures for the designated queue a current throughput of the incoming data packets and a current rate of interrupts, makes a comparison between the current measured throughput and rate of interrupts to the throughput and rate of interrupts that were measured during a preceding period in the succession, and selects and applies an update to the interrupt moderation parameter responsively to the comparison.

Dynamically-tuned interrupt moderation

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.

Remote host management over a network

A startup sequence in a computer system is initiated by detecting a bus reset event in an I/O device connected to a host, and responsively to the bus reset event communicating resources required to be allocated by the host. When a startup command from a host driver is not received within a predetermined bus reset count, the device autonomously changes its current configuration to a safe mode configuration, wherein fewer resources are required to be allocated relative to the current configuration. The safe mode configuration is communicated from the device to the host.

Hardware device safe mode

Growing community of inventors

Moran, Israel

Amir Ancel