CN1645396A - Applying method for linux group system in securities line - Google Patents
Applying method for linux group system in securities line Download PDFInfo
- Publication number
- CN1645396A CN1645396A CN 200510009507 CN200510009507A CN1645396A CN 1645396 A CN1645396 A CN 1645396A CN 200510009507 CN200510009507 CN 200510009507 CN 200510009507 A CN200510009507 A CN 200510009507A CN 1645396 A CN1645396 A CN 1645396A
- Authority
- CN
- China
- Prior art keywords
- node
- index
- network
- workstation
- linux
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
A Linux computer group system has index calculation platform including storage region, production region, login region, backup test region and I/O region. The application method of the system includes setting up configuration of computer group; setting computer group network address to be two different sections; setting up devices of storage region, production region, backup test region, logging node, I/O region, monitoring node, exchanger and KVM.
Description
Technical field
The present invention relates to the Computer Applied Technology field, the application process of particularly a kind of Linux Network of Workstation in securities industry.
Background technology
Cluster server be one group separate, by the interconnected computing machine of express network, and managed with the pattern of triangular web, promptly make full use of the resource of each computing machine in the group of planes, realize the parallel processing of complex calculation.The characteristic of its high-performance, high available, easy expansion can better meet the demand of current growing information service.At present, the Network of Workstation server has been widely used in applications such as scientific research, meteorology, oil, electric power, biology, CAD.Group of planes technology has been the publicly-owned technology of comparative maturity.But the application of a group of planes is constantly expansion.
Security industry index calculate have mass data processing, real-time height, requirement that reliability is high.The real-time processing of index, issue, the calculating etc. of recalling of index requires index computing system to have good real time performance, stability and reliability.And relating to reading, handling of mass data in the Index for Calculation process, these all determine index computing system should be implemented on a kind of efficient, stabilised platform.Index system adopt the Unix minicomputer as construction platform, and all there is relatively more significant drawback in this mode on speed, stability and reliability still mostly based on the Unix minicomputer at present.Simultaneously, the software systems of developing based on this mode are strong to the hardware platform dependence, are difficult for transplanting.These restraining factors are all from having fettered industry development to a certain degree.
As the system platform that security industry index calculates, utilizing the cluster server system to carry out Index for Calculation the cluster server system, is a brand-new technology field.Make full use of the superiority of Network of Workstation aspect high-performance calculation, improve the performance of index system, can better promote the development of securities industry at computing, safety, aspect such as reliable.
The application Network of Workstation is carried out Index for Calculation and is had following advantage:
1. group of planes practical application performance obviously is better than Unix minicomputer system, adopts the parallelization treatment technology can better embody the superiority of Network of Workstation on performance; In the stock index based on the Network of Workstation of Linux calculates, adopt the parallelization Index for Calculation, utilize the data read of parallelization and distribution technique very large program improve Index for Calculation speed.
2. a group of planes has superior extensibility.Adopt group of planes technology, can be according to the needs dynamic expansion node of using, the gridding program design, user friendly secondary development, convenient, flexible.
3. high available group of planes technology can guarantee enough availabilities;
4. the high reliability of system.Adopt highly reliable Network of Workstation, eliminate Single Point of Faliure.Improve the reliability of system greatly.
5. the good stability of system.Network of Workstation adopts ripe product technology, has strong application support and technical support.
6. group of planes inside is provided with special memory device, preservation backup that can more effective realization significant data.
7. the Network of Workstation cost is relatively low;
The (SuSE) Linux OS development in recent years is rapid, and its security, reliability, stability and portability have attracted increasing attention.Linux has the most powerful development teams in the whole world as the operating system of increasing income, and makes Linux obtain good development these years, also becomes an outstanding operating system platform simultaneously.The development of Linux has obtained IBM, and the support energetically of international well-known manufacturers such as Oracle, HP, Intel, Bea, Dell makes Linux all have outstanding solution and successful case from hardware platform, database platform.Simultaneously, Linux does not rely on hardware platform, not only can stable operation on 32 and 64 s' PC server, on multiple minicomputer, also obtained good support simultaneously.Linux is the operating system that can support Anthem II, many kinds 64 bit CPUs of EMT64, AMD64 at present fully.
Two kinds of technology of a Linux and a group of planes combine, and can fundamentally solve security industry index and calculate many-sided requirements such as real-time, stable, reliable and portables.
Summary of the invention
Main contents of the present invention comprise the index computing system that utilizes the Linux Network of Workstation to make up securities industry.This Design Mode can better meet the requirement of Index for Calculation real-time, stability for security industry index computing system provides a kind of novel structure model.Index system based on its exploitation has higher real-time calculating and recalls calculated performance, better reliability, dirigibility and extensibility.
Stock index is meant selects representative security in numerous security, utilize the security exchange price, the weight capital stock, and base period and base period index etc. carry out COMPREHENSIVE CALCULATING and the data that obtain, it can react the situation of a class security exchange, has the scale function.After it was as a kind of product official listing transaction, it had just had the investment function again so.Calculate stock index, purpose is to pick out A representative Sample security by the objective method of science, set up the general picture of a reflection securities market and operation conditions, can be as the reference index on investment evaluation yardstick and financial derivative product basis.
Make the Linux Network of Workstation can be applied to this brand-new field of Index for Calculation in the securities industry.
The index computing system that utilizes the multinode Network of Workstation to make up adopts (SuSE) Linux OS as operating system platform, has safety, reliable, stable and transplantable characteristic.
The application process of a kind of Linux Network of Workstation in securities industry comprises: Network of Workstation is applied to Index for Calculation field in the securities industry, by being used, Index for Calculation is transplanted on the Network of Workstation, satisfy Index for Calculation to counting yield, computed reliability calculates computation requirements such as extensibility.
Index computing system is structured on the group of planes hardware platform, the inner a plurality of node cooperative works of a group of planes, the computing of fellowship index.
Group of planes inside comprises production area and memory block, and the production area is used for the index computing, and the memory block is used to carry out preservation, the backup of valid data.
The processing node of Network of Workstation is based on (SuSE) Linux OS.
On the inner a plurality of processing nodes of Network of Workstation parallel processing software is installed, the multiple spot parallel processing of index support computing system utilizes the parallelization technology to improve Index for Calculation speed.
The Index for Calculation platform can mainly be divided into according to function: memory block, production area, logon area, backup test site and I/O district.
The memory block mainly is made up of two nodes and a HP disk array.It is the storage area of system data, is mainly used in the mass data that the storage Index for Calculation needs, and the deposit data of system is had important effect, is the nucleus of system.The memory block is two mutually redundant machines, carries out backup management by HA software.Then adopt several different methods protection for significant data, comprise disk RAID (finishing automatically), power-off protection methods such as (realizing) by using ups power by the HP disk array.
The production area is made up of a plurality of nodes, is the main operational area of system, mainly finishes parallel computation work, guarantee that it can move in 24 hours.Use group of planes parallel computing to carry out Index for Calculation, a plurality of node cooperative works in the group of planes, parallel computation is finished the real-time calculating of index jointly and is recalled task such as calculating.The computing MPI program of operational system on the node, the parallel computation of index support system is handled, requirement is the computing of deployment system rapidly, and can fast operation result be aggregated into memory node, to guarantee also that simultaneously system has high reliability, if have node to go wrong in the operation node, system node can carry out balanced distribution to load automatically.
Logon area in the system comprises a node, is responsible for the login and the monitoring of system, comprises user's log-on message in the file system of this node, and user's management and control also are configured on this login node.Simultaneously, systematic supervisory system is installed on this node, finishes the monitoring function of system.
The backup test site is made up of 8 nodes, forms the backup test site of system, is responsible for standby backup and test.When go wrong in the system node or node load surpassed the door threshold value of default, backup node just dropped into actual production.
The I/O district is made up of a node, is responsible for the outbound data issue of system.
Description of drawings
Fig. 1 utilizes Network of Workstation to dispose the Index for Calculation synoptic diagram.
Fig. 2 is the application process process flow diagram of Linux Network of Workstation in securities industry.
Embodiment
The Network of Workstation of utilizing of Fig. 1 is disposed Index for Calculation, and Network of Workstation is disposed the difference in functionality node that Index for Calculation need dispose Network of Workstation a group of planes according to demand.The data importing node is responsible for the input and the reception of system data, computing node is responsible for Index for Calculation, the Web service node is responsible for wide area network is carried out the index issuing service, backup problem when backup node is responsible for system and is broken down, the login node is responsible for the login management of system, memory node is responsible for the data storage of system, and test node carries out index for the user test is set.Connect with dual network between a group of planes, guarantee the reliability of grid.
The application process of Linux Network of Workstation in securities industry of Fig. 2, its step is as follows:
Step S1 is configured installation according to the characteristics of a group of planes to Network of Workstation, needs the standard cabinet of 600 (cm) * 900 (cm) * 2000 (cm) (length and width height); Demand has the rack of good performance of ventilating.Requirement can provide the rack of strong fan; Every node all needs a dividing plate to isolate.Require the thickness of dividing plate can not surpass 1cm; Keep good performance of ventilating, can guarantee that the temperature of machine does not surpass 45 degree; The power of every machine is 350W, amounts to 350*24=8400W; The power supply wiring board of 30 sockets can be provided; Require to be connected with 6 class lines between gigabit switch and server, 100 m switch requires 5 class lines to connect.Recommending all to carry out network with six class lines connects.
Step S2, the network address that a group of planes is set is two different network segments; For example: 10.10.10.0/255.255.255.0 and 192.168.1.0/255.255.255.0.Can determine two inner network segments according to internal networking structure.
Step S3, install the memory block, and the memory block is made up of two PC servers and storage array.Two servers are connected to each other by the level on line.One as machine for producing, order outer one as the backup machine.Two-server can be connected on the storage array of HP simultaneously.
Server configures: 2 of CPU:Intel Xeon2.8G/D L2:512k
2 of 1G Registered ECC DDR
Seagate hard disk 36GB 10000 changes 1 of Ultra320 hard disk
1 of CD-ROM
1.44 " 1 of floppy drive
1 of gigabit RJ45 interface adaptive network interface card
1 of 100,000,000 RJ45 interface adaptive network interface card
Install software: Redhate9.0 Kernel linux-2.4.20-8
Oracle?9i?Enterprise?Standard?Edition
MPI concurrent software mpich-1.2.5.2
Step S4 generates the district and installs, and the production area has the node machine of 12 actual productions to form, and finishes the calculating of system.
Server configures: identical with memory area configuration.
Install software: Redhate9.0 Kernel linux-2.4.20-8
MPI concurrent software mpich-1.2.5.2
Step S5, the backup test site is installed, and there are 8 nodes the backup test block, backs up the index new with pilot development for the user.
Server configures: trivial configuration is identical with producing.
Install software: identical with the production area configuration.
Step S6, the login node is installed, and login is made up of a node, is used for the login and the monitoring of system.
Server configures: trivial configuration is identical with producing.
Install software: Redhat9.0 Kernel linux-2.4.20-8
Executed in parallel command software Prsh
Step S7, I/O district install, and I/O district is made up of a node, be used for system's outbound data and issue, and with being connected of other system.
Server configures: trivial configuration is identical with producing.
Install software: Redhat9.0 Kernel linux-2.4.20-8
Index for Calculation research/development platform environment
Step S8, monitor node is installed,
Monitor node install software: Windows2000Professional SP4
ServerMaster
Step S9, switch is installed
Equipment, Catalyst 3750 Switch (gigabit switch): 24 mouthfuls, the 10/100/1000Mbps self-adaptation.
Equipment, Catalyst 2950 Switch (100 m switch): 48 mouthfuls, the 10/100Mbps self-adaptation.
Two switches should be installed on the most significant end of rack.Connect with different server by netting twine.
Step S10, KVM install and are equipped with Raritan 24 Z-Series Z4200U
Link control desk by netting twine.At control desk a group of planes is managed.
Claims (8)
1. the application process of a Linux Network of Workstation in securities industry comprises: Network of Workstation is applied to Index for Calculation field in the securities industry, by being used, Index for Calculation is transplanted on the Network of Workstation, satisfy Index for Calculation to counting yield, computed reliability calculates computation requirements such as extensibility.
2. system according to claim 1 is characterized in that: index computing system is structured on the group of planes hardware platform, the inner a plurality of node cooperative works of a group of planes, the computing of fellowship index.
3. system according to claim 1 is characterized in that: group of planes inside comprises production area and memory block, and the production area is used for the index computing, and the memory block is used to carry out preservation, the backup of valid data.
4. system according to claim 1 is characterized in that: the processing node of Network of Workstation is based on (SuSE) Linux OS.
5 systems according to claim 1 is characterized in that: on the inner a plurality of processing nodes of Network of Workstation parallel processing software is installed, the multiple spot parallel processing of index support computing system utilizes the parallelization technology to improve Index for Calculation speed.
6. the application process of a Linux Network of Workstation in securities industry, Network of Workstation adopts the (SuSE) Linux OS platform, be divided into: memory block, production area, logon area, backup test site and I/O district, the memory block mainly is made up of two nodes and a HP disk array, be the storage area of system data, be used to store the mass data of Index for Calculation needs;
The production area is made up of a plurality of nodes, is the operational area of system, finishes parallel computation work, the computing MPI program of operational system on the node, the parallel computation of index support system is handled, and requirement is the computing of deployment system rapidly, and can fast operation result be aggregated into memory node;
Logon area comprises a node, is responsible for the login and the monitoring of system, comprises user's log-on message in the file system of this node, and user's management and control also are configured on this login node;
The backup test site is made up of 8 nodes, forms the backup test site of system, is responsible for standby backup and test, and when go wrong in the system node or node load surpassed the door threshold value of default, backup node just dropped into actual production;
The I/O district is made up of a node, is responsible for the outbound data issue of system.
7. according to the application process of Linux Network of Workstation in securities industry of claim 6, it is characterized in that the memory block is two mutually redundant machines, carries out backup management by HA software.
8. according to the application process of Linux Network of Workstation in securities industry of claim 6, its concrete steps are as follows;
Step S1 is configured installation according to the characteristics of a group of planes to Network of Workstation;
Step S2, the network address that a group of planes is set is two different network segments;
Step S3, install the memory block, and the memory block is made up of two PC servers and storage array, and two servers are connected to each other by the level on line;
Step S4 generates the district and installs;
Step S5, the backup test site is installed;
Step S6, the login node is installed;
Step S7, install in the I/O district;
Step S8, monitor node is installed;
Step S9, switch is installed;
Step S10, KVM installs;
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 200510009507 CN1645396A (en) | 2005-02-22 | 2005-02-22 | Applying method for linux group system in securities line |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 200510009507 CN1645396A (en) | 2005-02-22 | 2005-02-22 | Applying method for linux group system in securities line |
Publications (1)
Publication Number | Publication Date |
---|---|
CN1645396A true CN1645396A (en) | 2005-07-27 |
Family
ID=34875382
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 200510009507 Pending CN1645396A (en) | 2005-02-22 | 2005-02-22 | Applying method for linux group system in securities line |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN1645396A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101888414A (en) * | 2010-07-02 | 2010-11-17 | 武汉大学 | Mobile multitask parallel processing system |
CN105610611A (en) * | 2015-12-23 | 2016-05-25 | 浪潮集团有限公司 | Method for remotely and automatically installing distributed big data system |
-
2005
- 2005-02-22 CN CN 200510009507 patent/CN1645396A/en active Pending
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101888414A (en) * | 2010-07-02 | 2010-11-17 | 武汉大学 | Mobile multitask parallel processing system |
CN101888414B (en) * | 2010-07-02 | 2013-02-13 | 武汉大学 | Mobile multitask parallel processing system |
CN105610611A (en) * | 2015-12-23 | 2016-05-25 | 浪潮集团有限公司 | Method for remotely and automatically installing distributed big data system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8006061B1 (en) | Data migration between multiple tiers in a storage system using pivot tables | |
Tiwari et al. | Active Flash: Towards {Energy-Efficient},{In-Situ} Data Analytics on {Extreme-Scale} Machines | |
Alam et al. | Parallel I/O and the metadata wall | |
KR20180027326A (en) | Efficient data caching management in scalable multi-stage data processing systems | |
CN111210879B (en) | Hierarchical storage optimization method for super-large-scale drug data | |
US20080028107A1 (en) | System and method for automatic reassignment of shared storage on blade replacement | |
Li et al. | An in-depth analysis of cloud block storage workloads in large-scale production | |
CN105677486A (en) | Data parallel processing method and system | |
CN101662495A (en) | Backup method, master server, backup servers and backup system | |
WO2010024027A1 (en) | Virtual server system and physical server selection method | |
US8024542B1 (en) | Allocating background workflows in a data storage system using historical data | |
CN102360327A (en) | Hard disk status information collection method and device | |
Xie et al. | Dynamic data reallocation in hybrid disk arrays | |
US11726856B2 (en) | Systems and methods for identification of issue resolutions using collaborative filtering | |
US11314431B2 (en) | Distributed data blocks using storage path cost values | |
US10437477B2 (en) | System and method to detect storage controller workloads and to dynamically split a backplane | |
Barroso et al. | The datacenter as a computer | |
US20110167229A1 (en) | Balanced data-intensive computing | |
CN1959643A (en) | Apparatus, system, and method for reassigning a client | |
CN1829230A (en) | Intelligent network disc storage system and its realizing method | |
CN1645396A (en) | Applying method for linux group system in securities line | |
Wei et al. | DSC: Dynamic stripe construction for asynchronous encoding in clustered file system | |
US7543089B2 (en) | Adaptive input/output bus sharing method to improve performance of SCSI/SAS clusters | |
CN116074179A (en) | High expansion node system based on CPU-NPU cooperation and training method | |
CN214011980U (en) | Server with RAS (remote server system) characteristic |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20050727 |