Tsinghua University

Other

Weimin Zheng

Wenguang Chen

Kang Chen

Jidong Zhai

Wei Xue

Hongliang Yu

Guangyan Zhang

Teng Ge Ma

Yongwei Wu

Mingxing Zhang

Maomeng Su

Yuyang Jin

A computer executed parallel program scalability bottleneck detection method is provided, which includes: building a program structure graph for a program source code; collecting performance data based on a sampling technique during runtime; the performance data including: performance data of each vertex of the program structure graph and inter-process communication dependence of communication vertices; building a program performance graph by filling the program structure graph with the collected performance data, the program performance graph recording data and control dependence of each process as well as inter-process communication dependence; detecting problematic vertices from the program performance graph, and starting from some or all of the problematic vertices, backtracking through data/control dependence edges within a process and communication dependence edges between different processes, to detect scalability bottleneck vertices.

Parallel program scalability bottleneck detection method and computing device

A hybrid network system, a communication method and a network node are provided. The hybrid network system includes a combined network system architecture of a point-to-point network and a broadcast network, the point-to-point network is provided with a plurality of point-to-point channels, and each network node can send a message to a corresponding network node or receive the message from the corresponding network node through the point-to-point channel; the broadcast network is provided with a plurality of broadcast channels, and each broadcast channel can only have one sender at most at the same time, but can have a plurality of receivers; at least a part of network nodes have the capability of sending messages to be broadcast, at least a part of network nodes have the capability of receiving the broadcast messages, and at least some network nodes are configured to select the broadcast network or the point-to-point network at least according to communication tasks to be executed per se to complete the communication tasks to be executed. Preferably, each network node has the capability of sending and receiving the broadcast message. The hybrid network system of the embodiment of the present invention has high flexibility and high communication performance.

Hybrid network system, communication method and network node

The data storage system is a RAID-based data storage system in which resources are globally shared. This storage system includes the first number of disks, and the RAID mechanism is used to store data on each disk. The blocks on different disks form stripes, and at least one of the blocks on the stripe stores the parity information, wherein the width of the stripe is less than the first number. The data layout of the data storage system satisfies the following characteristics: any two physical blocks in the stripe are distributed on different disks; the data blocks distributed on each disk are the same, and the distributed parity blocks are also the same; other data in the stripe associated with any piece of disk data is evenly distributed across all the remaining disks. Normal data layout and degraded data layout can be implemented by orthogonal Latin squares. This system can remove the limitation that the number of disks in the normal data storage system is equal to the stripe width, and break the resource isolation between the disk groups. And in the event of a disk failure, this invention can achieve a complete equalization of the reconstructed read load.

RAID-based globally resource-shared data storage system

Provided are a method for multiple accesses in a memory, an apparatus supporting multiple accesses in the memory and a storage system. The method comprises: receiving a number N of addresses in the memory, wherein N is an integer larger than 1 and the number N of addresses are discontinuous; performing a preset operation according to the number N of addresses; and outputting the result of the operation. As a consequence, the performances of a computer system can be improved, and a user can input and use the desired addresses just as required by the user.

Method and apparatus for multiple accesses in memory and storage system, wherein the memory return addresses of vertexes that have not been traversed

Systems and methods are provided for performing a remote procedure call. One method may comprise, at a client device, generating a request including setting a status field in a request header to indicate to a server processor that the request is ready, writing the request to a server memory via a RDMA write operation and fetching a response generated by the server processor from the server memory via a RDMA read operation. The method may further comprise, at a server device, checking a mode flag to determine that an operation mode is set to repeated remote fetching, retrieving the request from a server memory, processing the request to generate a response and writing the response to the server memory for the response to be fetched by a client device. The response includes a response header that comprises a status field for the status of the response and a response time.

Systems and methods for remote procedure call

The data storage system and the data storage approach based on redundant arrays of independent disks (RAID) is provided. The storage system includes: an RAID controller, and a storage device subsystem. This storage device subsystem includes a plurality of first physical storage devices for data storage and a plurality of second physical storage devices for parity storage. First physical storage devices are divided into a plurality of sets of first disks which are independent from each other. The data accesses to different sets of first disks will not interfere with each other. These second disks are shared by all the sets of first disks. The RAID controller stores data in each set of first disks and stores parity information in multiple shared second disks by the RAID mechanism. The storage system combines the parity from different RAID volumes into one shared set and this improves the performance and reliability of the whole storage system.

Data storage system and data storage method

Embodiments disclosed herein related to a virtual machine-based on-demand parallel disaster recovery system and a method thereof. By integrating context of processes of virtual machine, the system and method build a process tree related to applications and support the operation of freezing the context and intercept memory image consistent with the virtual machine environment, based on which, a method for on-demand parallel recovery is provided. In the method, while monitoring file system data that processes of virtual machine are to visit by a kernel intercept unit, recovery of related data blocks of the applications is conducted in the background in order to make applications and recovery procedure of the virtual machine run simultaneously. The recovery procedure is divided into two phases, recovering memory image first and then recovering file system or volume data. Through the procedure, applications can be recovered precisely to the running state when they are backup. The method realizes the scheduling of multiple backup tasks and recovery tasks in client and realizes the prefetching of the recovered data. The embodiments described herein consume less time for recovery, enhances the reliability of the protected applications, has a low cost and practical value; and it is transparent to users.

Virtual machine-based on-demand parallel disaster recovery system and the method thereof

Embodiments disclosed herein related to a system used for disaster recovery backup. The system converts the operating system in the system preparing for disaster recovery into a virtualized system which will be disaster-recovery-ready. The system includes: a device to be converted on which a physical operating system is installed; a USB converting device which is coupled to the device to be converted, which includes a USB disc operating system, a converting unit and a virtual system, and which is used to convert the physical operating system in the device to be converted into a virtual system for making backup of the virtual system. Embodiments disclosed herein are further related to a method for disaster recovery backup and a method for installing the disaster recovery system. The disaster recovery system based on the virtualization technology may be deployed rapidly without making any changes to the physical operating system itself. After conversion, applications in the physical operating system are not influenced by the process of disaster recovery. In addition, because the running environment of virtual system after conversion and that of original physical operating system share the same physical machine, there is no unwanted influence on the virtual system after conversion caused by the difference of properties.

Growing community of inventors

Beijing, China

Weimin Zheng