Hewlett Packard Enterprise Development LP

Compaq Computer Corporation, Inc.

Hewlett-Packard Development Company, L.p.

John L Byrne

Harumi Kuno

Bruce J Walker

Laura L Ramirez

Parthasarathy Ranganathan

Wei Zhang

Mark David Lillibridge

Duncan Roweth

Alan Lynn Davis

Sai Rahul Chalamalasetti

Sergey Serebryakov

Torsten Wilde

Ryan Dean Menhusen

Daniel William Dauwe

William W Chow

David B Zafman

Jeffrey Drummond

Khemraj Shukla

Vipin Kumar Kukkala

John A Gertwagen

One embodiment provides a system and method for predicting network power usage associated with workloads. During operation, the system configures a simulator to simulate operations of a plurality of network components, which comprises embedding one or more event counters in each simulated network component. A respective event counter is configured to count a number of network-power-related events. The system collects, based on values of the event counters, network-power-related performance data associated with one or more sample workloads applied to the simulator; and trains a machine-learning model with the collected network-power-related performance data and characteristics of the sample workloads as training data 1, thereby facilitating prediction of network-power-related performance associated with a to-be-evaluated workload.

System and method for predicting power usage of network components

Methods and systems for conducting vector send operations are provided. The processor of a sender node receives a request to perform a collective send operation (e.g., MPI_Broadcast) from a user application, requesting a copy of data in one or more send buffers by sent to each of a plurality of destinations in a destination vector. The processor invokes a vector send operation from a software communications library, placing a remote enqueue atomic send command for each destination node of the destination vector in an entry of a transmit data mover (XDM) command queue in a single call. The processor executes all of the commands in the XDM command queue and writes the data in the one or more send buffers into each receive queue of each destination identified in the destination vector.

Vector send operation for message-based communication

A system includes a plurality of processes, a network fabric, and a shared memory accessible by the plurality of processes over the network fabric, the shared memory to store a plurality of elements of a data structure. A first process is designated as being allowed to update a target variable stored in the shared memory, and a second process of the plurality of processes writes a request for an atomic operation to a first region in the shared memory. The first process is responsive to the request to perform the atomic operation that updates the target variable, and write a result including a value of the updated target variable to a second region in the shared memory, the second region readable by the second process, the request and the result being elements of the data structure.

Atomic operations for fabric shared memories

Techniques for memory store error checks are provided. In one aspect, a process running on a processor may execute an instruction to store a first value in memory. The processor may store a plurality of values, including the first value, from a plurality of processes to the memory. A check on a synchronous error notification path may be performed to determine whether an error in storing at least one of the plurality of values occurred.

Memory store error check

Illustrated is a system and method for executing a checkpoint scheme as part of processing a workload using an application. The system and method also includes identifying a checkpoint event that requires an additional checkpoint scheme. The system and method includes retrieving checkpoint data associated with the checkpoint event. It also includes building a checkpoint model based upon the checkpoint data. The system and method further includes identifying the additional checkpoint scheme, based upon the checkpoint model, the additional checkpoint scheme to be executed as part of the processing of the workload using the application.

Dynamically modeling and selecting a checkpoint scheme based upon an application workload

A method and apparatus for transparent failover of a filesystem within a computer cluster is provided. For failover protection, a filesystem is physically connected to an active server node and a standby server node. A cluster file system provides distributed access to the filesystem throughout the computer cluster. The cluster file system monitors the progress of each operation performed on the failover protected filesystem. If the active server node should fail during an operation, all processes performing operations on the failover protected filesystem are caused to sleep. The filesystem is then relocated to the standby server node. The cluster file system then awakens each sleeping process and retries each pending operation.

Root filesystem failover in a single system image environment

Filesystem failover in a single system image environment

Growing community of inventors

Pasadena, CA, United States of America

John L Byrne