CN104394223A - Automatic rapid deployment method of large-scale computer cluster system nodes - Google Patents
Automatic rapid deployment method of large-scale computer cluster system nodes Download PDFInfo
- Publication number
- CN104394223A CN104394223A CN201410718742.8A CN201410718742A CN104394223A CN 104394223 A CN104394223 A CN 104394223A CN 201410718742 A CN201410718742 A CN 201410718742A CN 104394223 A CN104394223 A CN 104394223A
- Authority
- CN
- China
- Prior art keywords
- node
- reflection
- destination node
- management
- destination
- 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
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000009434 installation Methods 0.000 claims abstract description 37
- 238000012360 testing method Methods 0.000 claims description 14
- 230000005540 biological transmission Effects 0.000 claims description 12
- KKIMDKMETPPURN-UHFFFAOYSA-N 1-(3-(trifluoromethyl)phenyl)piperazine Chemical compound FC(F)(F)C1=CC=CC(N2CCNCC2)=C1 KKIMDKMETPPURN-UHFFFAOYSA-N 0.000 claims description 11
- 238000012546 transfer Methods 0.000 claims description 6
- 230000000977 initiatory effect Effects 0.000 claims description 5
- 238000013507 mapping Methods 0.000 abstract 6
- 230000002093 peripheral effect Effects 0.000 abstract 1
- 230000008569 process Effects 0.000 description 6
- 238000013515 script Methods 0.000 description 6
- 238000005192 partition Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 239000003795 chemical substances by application Substances 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000011900 installation process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000010076 replication Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000008521 reorganization Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0876—Aspects of the degree of configuration automation
- H04L41/0886—Fully automatic configuration
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
Abstract
The invention provides an automatic rapid deployment method of large-scale computer cluster system nodes. Based on a client machine/server system structure, node software installation is based on mapping, the mapping is an overall system which comprises an operating system and an application software packet which need to be installed on a target node, and all mappings are stored in a server. The method disclosed by the invention provides a simplex control point to a user; all control and operation can be finished in a managing node as long as the user is connected to the managing node through a terminal. The target node needs no peripheral equipment such as a keyboard, a mouse, a displayer, a CD driver and a floppy driver; a plurality of nodes and different mappings can be installed simultaneously. By binding the mappings with the nodes, the corresponding relationship between the mappings and the nodes can e determined; a plurality of nodes can be installed simultaneously without mutual interference; the operating system and the application software packet of the cluster system nodes are full-automatic in installation and configuration; the target nodes need no manual intervention.
Description
Technical field
The present invention relates to computer cluster technology field, specifically a kind of automated quick deployment method of large-scale computer cluster node.
Background technology
High-Performance Computing Cluster computing system is applied widely at present, and the scale of cluster is also in continuous expansion, and at structure and the run duration of cluster, in cluster great deal of nodes, the artificial repetition installation and upgrade of operating system and application software package can become a difficult task.
In prior art before the present invention, generally with software for copying hard disc that the is manual or exploitation of ghost(Symantec company) hard disk imprinting or NFS(net file system, NFS) shared network installation method comes deployment operation system and application software.Manual methods, in CD mode, install time and the manpower of each node meeting at substantial, and the layoutprocedure of node is very loaded down with trivial details, is easy to make mistakes.Method conventional in the engineering practice building cluster is the method for ghost hard disk imprinting, and this method needs to dismantle the hard disk of mounted node, and requires that target disk is identical with source disk size; Or NFS shared network Method for Installation, namely using the node of in network as nfs server, the file copy in operating system installation dish on this node, mounted node is started by network interface card or floppy disk, completes installation from server downloading files.
Summary of the invention
The object of the invention is to overcome the deficiencies in the prior art, a kind of automated quick deployment method of large-scale computer cluster node is provided, to simplify cluster installment work, raise the efficiency, realize the full automation of installation process.
Technical scheme of the present invention realizes in the following manner, based on client/server architecture in its structure, node software installs the mode adopted based on reflection, and reflection is a complete system, containing the operating system needing to install on the target node and application software package; All reflections all store on the server; A management node is provided with as server in hardware structure needed for this method, destination node is connected with management node by commodity ethernet switch, destination node needs the network interface card and mainboard of supporting Pre-Boot eXecution Environment and WOL, destination node is connected with management node by commodity ethernet switch as the client computer in this structure;
1), management node is the server end in architecture, comprise the service end of server end for image transmission and Pre-Boot eXecution Environment agreement, be responsible for creating reflection, management reflection, store destination node information, WOL destination node and the various services required for deployment being provided;
2), destination node is client in architecture, obtains and disposes agency, dispose agency and be responsible for downloading from the management node as server video and being installed on the hard disk of this locality from management node;
3), the start-up course of destination node observes the Pre-Boot eXecution Environment agreement of Intel, management node provides dynamic host configuration protocol DHCP, simple file transfer protocol TFTP serve and service for transmitting reflection;
4), between destination node and management node, carry out image transmission in, use the different remote copy instruments based on client/server architecture;
The concrete steps of cluster node automated quick deployment method are:
Step 1: obtain reflection:
Reflection is obtained by following two kinds of modes:
The first creates reflection by installation kit, namely the installation kit of operating system in fixing disc and application software copied in a catalogue of management node, directly on management node, create virtual mapping with installation kit;
The second is from a source node snapping reflection, namely what a node of first manual installation, as the source node of reflection, source node is installed the various application software of operating system and needs, then catch the reflection of source node from management node and store, applicable destination node and source node are the situations of isomorphism;
Step 2: binding reflection and destination node:
The network identification information of destination node and reflection binds, determines to video and the corresponding relation of destination node, one of them corresponding multiple destination node of videoing, but destination node can only the unique reflection of correspondence;
Step 3: installation targets node:
When disposing, first on management node, start DHCP service, for destination node provides network identification information; Start TFTP service, download deployment Agent for destination node and service is provided; Be enabled in the server end of the remote copy instrument transmitting reflection between management node and destination node;
Management node starts destination node by Wakeup On LAN; After destination node is started by network interface card, first obtain network identity according to dynamic host configuration protocol DHCP from management node, then download according to simple file transfer protocol TFTP and dispose agency and run in local memory; Dispose agency and download reflection from management node, complete installation in this locality, after installation, node autoboot;
Step 4: configuration destination node:
When destination node installs initiating switchup, run configurator and the operating system of destination node and application software are configured; After configuration, namely destination node becomes a clustered node used immediately;
Step 5: test:
The all node installation of group system is complete, and testing results program, whether the Install and configure of Test Operating System and application software is correct.
Described destination node is placed through network interface card and starts in BIOS, its software installation process full automation; Destination node does not need the ancillary equipment such as keyboard, mouse, display, CD-ROM drive, floppy drive; All operations all complete on management node.
Advantage of the present invention is:
A, provide single control point to user, user only needs to be connected on management node by terminal, and all control operations all complete on management node.Destination node only needs in BIOS, be placed through network interface card and starts, and its software installation process is full automatic; Destination node does not need the ancillary equipment such as keyboard, mouse, display, CD-ROM drive, floppy drive.
B, multiple node and different reflections can be installed simultaneously.By the binding between reflection and node, determine reflection
And the corresponding relation between node, multiple node can be installed simultaneously, can not disturb each other.
C, cluster node operating system and application software package Install and configure full automation, destination node does not need completely
Manual intervention.In batches node can be installed by this method, be suitable for large-scale cluster system is installed.
Execution mode
Below the automated quick deployment method of large-scale computer cluster node of the present invention is described in detail below.
The automated quick deployment method of large-scale computer cluster node of the present invention, based on client/server architecture in its structure, node software installs the mode adopted based on reflection, reflection is a complete system, containing the operating system needing to install on the target node and application software package; All reflections all store on the server; A management node is provided with as server in hardware structure needed for this method, destination node is connected with management node by commodity ethernet switch, destination node needs the network interface card and mainboard of supporting Pre-Boot eXecution Environment and WOL, destination node is connected with management node by commodity ethernet switch as the client computer in this structure;
1), management node is the server end in architecture, comprise the service end of server end for image transmission and Pre-Boot eXecution Environment agreement, be responsible for creating reflection, management reflection, store destination node information, WOL destination node and the various services required for deployment being provided;
2), destination node is client in architecture, obtains and disposes agency, dispose agency and be responsible for downloading from the management node as server video and being installed on the hard disk of this locality from management node;
3), the start-up course of destination node observes the Pre-Boot eXecution Environment agreement of Intel, management node provides dynamic host configuration protocol DHCP, simple file transfer protocol TFTP serve and service for transmitting reflection;
4), between destination node and management node, carry out image transmission in, use the different remote copy instruments based on client/server architecture;
The concrete steps of cluster node automated quick deployment method are:
Step 1: obtain reflection:
Reflection is obtained by following two kinds of modes:
The first creates reflection by installation kit, namely the installation kit of operating system in fixing disc and application software copied in a catalogue of management node, directly on management node, create virtual mapping with installation kit;
The second is from a source node snapping reflection, namely what a node of first manual installation, as the source node of reflection, source node is installed the various application software of operating system and needs, then catch the reflection of source node from management node and store, applicable destination node and source node are the situations of isomorphism;
Step 2: binding reflection and destination node:
The network identification information of destination node and reflection binds, determines to video and the corresponding relation of destination node, one of them corresponding multiple destination node of videoing, but destination node can only the unique reflection of correspondence;
Step 3: installation targets node:
When disposing, first on management node, start DHCP service, for destination node provides network identification information; Start TFTP service, download deployment Agent for destination node and service is provided; Be enabled in the server end of the remote copy instrument transmitting reflection between management node and destination node;
Management node starts destination node by Wakeup On LAN; After destination node is started by network interface card, first obtain network identity according to dynamic host configuration protocol DHCP from management node, then download according to simple file transfer protocol TFTP and dispose agency and run in local memory; Dispose agency and download reflection from management node, complete installation in this locality, after installation, node autoboot;
Step 4: configuration destination node:
When destination node installs initiating switchup, run configurator and the operating system of destination node and application software are configured; After configuration, namely destination node becomes a clustered node used immediately;
Step 5: test:
The all node installation of group system is complete, and testing results program, whether the Install and configure of Test Operating System and application software is correct.
Described destination node is placed through network interface card and starts in BIOS, its software installation process full automation; Destination node does not need the ancillary equipment such as keyboard, mouse, display, CD-ROM drive, floppy drive; All operations all complete on management node.
Described deployment agency adopts and manually customizes, and it is a perfect micro operation system, runs in target node memory, independent of destination node hard disk; The instrument dispose the client including remote copy instrument in agency, necessary hard disk and NIC driver, hard disk being operated.
This method is when concrete enforcement, management node and destination node can select common business computer, management node requires there are the input-output equipment such as keyboard, mouse, CD-ROM drive, floppy drive and monitor, and destination node does not then need these input-output equipment.Management node requires larger hard disk (being greater than 80G) and is used for memory map.Require that the network interface card of destination node can support PXE and Wakeup On LAN, only in this way could from startup of server destination node, whole automations of realize target node installation process.Business computer hardware general at present can support this two functions.
Connected by Ethernet between destination node and management node, because image file is usually larger, so the efficiency impact of the network bandwidth on this method is very large, requiring that the bandwidth of Ethernet is is at least 100Mbps, and network medium can be twisted-pair feeder and optical fiber.Connected mode can pass through 100M or 1000M universal exchange.Adopt star-like or point-to-point network model, all destination nodes directly can be communicated with management node.
When transmitting reflection between management node and destination node, need the remote copy instrument that selection one is suitable, these Replication Tools can be selected according to actual conditions, and the operating system with reflection is relevant, and these Replication Tools need to carry out Read-write Catrol to the file system of reflection.Such as, in (SuSE) Linux OS, the instrument of software rsync as remote copy of remote synchronization can be selected.
In the method, dispose the micro operation system that agency is a perfect in shape and function, dispose the instrument that agency comprises the client of remote copy instrument, necessary hard disk and NIC driver, operates hard disk.When concrete enforcement, can be obtained by customization Linux.Dispose agency to need to run in the internal memory of destination node, and can identify the machine hard disk and by network and server communication, so need, according to the different hardware of destination node, to compile new linux kernel, add the driver of hardware.Make the root file system corresponding with new kernel to video RAMDISK file, add wherein necessary hard disk is operated instrument, remote copy instrument (such as rsync) client, driver module (such as SCSI driver module) etc.
Concrete steps are as follows:
Step 1: obtain reflection
The mode obtaining reflection has two kinds, and one creates reflection on management node, and another kind catches reflection from certain source node, introduces the concrete steps of these two kinds of modes below respectively.
Creating reflection is exactly use some resources to generate reflection on management node, and these resources comprise operating system nucleus, application software package, disk partition file etc., also needs, according to doing some configurations to reflection, that is to write the script needing some configurations.Then these resources are carried out installation reorganization and conformity according to certain rule, form a relatively independent system, this system is exactly a reflection.
The step creating reflection is generally as follows:
1) reflection stake is knocked in
Stake is the basis of this reflection, and this needs the structure that creates a directory, and then loads the file system of some indispensabilities of current system.To some degree, creating reflection is copying management node operating system, but just replicates little file of management node operating system itself, later establishment is had " stake ".
2) installation system bag
After creating " stake ", just can, based on " stake ", namely root be changed in stake, the bag that installation system needs, become one and have nothing to do with original file system, relatively independent file system, this have just had the prototype of image file system.
3) configuration image
Although second step has had reflection, but this reflection can't use, and also needs to carry out some configurations to this reflection, and as disk partition etc., such reflection could use.Certainly, this configuration can be changed, and the reflection based on file mode provides greater flexibility.
The process catching reflection is similar to the working method of taking pictures, first manual installation source node, installing operating system and necessary application software package on the source node, then a reflection of distortionless acquisition source node management node, in fact acquisition process is exactly a process copied, and the remote copy instruments such as rsync can be used.The step catching reflection is as follows:
1) system is remained static to be convenient to obtain effective snapshot
The operating system run needs to safeguard its operating state, the data of all in store operating state in internal memory and in hard disk.Disk snapshot can not clone the state of hard-disk system completely.Therefore, when carrying out reflection seizure to source node system, operating system must do the smooth enforcement that reflection capture-process is convenient in suitable adjustment.Such as, sometimes need to delete some identifying informations, as journal file, host name, IP address etc. temporarily.The reflection obtained so just can be made to be the file of a phase centering, to be conveniently transplanted in other nodes.
2) a distortionless reflection is caught
First want detection resources node whether can catch reflection, then start image catches agency, reads the file in source node disk, they are stored in the hard disk of management node.
3) configuration image
Necessary configuration is carried out to the reflection obtained, different from establishment image mode, also source node while catching reflection
Disk partition information reproduction come, therefore do not need configure disk partition.Here configuration is mainly to the configuration of application software and operating system in reflection.
Step 2: binding reflection and destination node
The network information such as IP address, host name, MAC Address of destination node and reflection are bound, determines the corresponding relation of reflection and destination node.Need to set up the corresponding relation that a toy data base comes memory node and reflection.Wherein, reflection and node are the relations of one-to-many, and namely reflection can for multiple node, and a node can only a corresponding reflection determined.
Step 3: installation targets node
When disposing, the service started is needed to comprise DHCP service, TFTP service, rsync service etc.Dynamic Host Configuration Protocol server provides IP address, the network informations such as host name for destination node, and specifies the position of tftp server, and notification target node goes to the position of specifying to download starting factor file.
Tftp server is downloaded deployment Agent for destination node and is provided service, stores starting factor file, starts configuration file, linux kernel and root file system etc. in the root of tftp server.Destination node downloads load and execution in these files to local memory.
All reflections are in charge of in rsync service, provide service for transmitting reflection between management node and destination node.On management node, rsync runs in the mode of finger daemon.After destination node starts, according to the reflection that the network identification informations such as host name own find the machine corresponding from server, then from management node, reflection is copied to local hard disk by rsync.
In this step, first need the configuration file generating DHCP service, configuration file generates according to network informations such as the destination node IP address stored in management node database, host name and MAC Address.Then on management node, start DHCP service, TFTP service and rsync service.
Destination node needs to connect with power supply, and in BIOS, be placed through network interface card startup.After management node starts respective services, send the broadcast packet comprising destination node MAC Address from management node, start destination node by WOL (Wakeup On LAN) technology.
Destination node is started by network interface card after receiving the broadcast packet corresponding with the machine MAC Address, first the network identification information corresponding with the machine is obtained from Dynamic Host Configuration Protocol server according to the MAC Address of the machine, then download network interface card Bootloader from tftp server and dispose agency, after deployment agency loads and starts in local memory, first detect and the hard disk of carry this locality, then according to disk partition configuration file, subregion is carried out to local hard drive.The remote copy instruments such as rsync are used to download the reflection corresponding with the machine from server and complete copy to local hard drive after subregion completes, after installation, node autoboot.
Step 4: configuration destination node
When destination node installs initiating switchup, run configurator and the operating system of destination node and application software are configured.Because group system be a loose coupling and multimachine system, in cluster, node is relatively independent, and each node has independently operating system and application software.There is provided server to external world in order to the node in cluster can mutually be write, just must be configured node.
In (SuSE) Linux OS, layoutprocedure performs a series of script exactly, and each application software has corresponding configuration script, and these scripts just write reflection and suffered in this step of acquisition reflection.Because the collocation method of often kind of application software is all different, thus configuration script to be write out according to concrete application software and stored in reflection in.When node initiating switchup, call these scripts, complete the configuration to application software.For Windows operating system, the configuration of software needs to call some configuration tools, realizes by the mode of batch processing.
After configuration, destination node has been exactly a clustered node that can use immediately.
Step 5: test: all node installation of cluster is complete, and testing results program, whether the Install and configure of Test Operating System and application software is correct.Test program is when being configured reflection in write reflection, it is the test of the basic function to application software, such as, parallel computation software MPICH. under linux system is after completing the configuration to software, can run a small routine write with MPI, whether inspection operation result is correct.
This test is closely-related with application, needs to carry out design test program according to different application programs.Because reflection can configure, test program can write in reflection in advance.
Due to this method employing is based on client/server (Client/Server) structure, and reflection is all stored on management node, and what all destination nodes were parallel downloads reflection from management node.So under the prerequisite of transmitting hypothesis and Fairshare bandwidth, time required for clustered deploy(ment) can be calculated according to the network bandwidth.The formula of calculating deployment time below can be drawn according to clustered deploy(ment) process:
Deployment time=transmission reflection time+the transmission reflection time
=transmission reflection time+node number × image size/network bandwidth
With letter representation be:
t=
c+n×
s×
8/W(formula 1)
Wherein T is deployment time;
cfor transmission reflection time, namely start to image deployment agency load operating internal memory from destination node network interface card and during this period of time, a constant can be thought;
nfor concurrently deployed destination node number;
sfor image size (MB);
wfor the network bandwidth (Mb/s).
In the cluster building process of Shandong Province's HLRS, employ the clustered deploy(ment) software according to the inventive method design, the hardware environment of this cluster is as follows:
Reflection obtains and adopts reflection catching mode, and the image size of acquisition is 3.7G.Being started shooting (having 9 nodes to be not activated successfully because of arranging reason) by WOL 87 nodes simultaneously, starting from network interface card and carrying out Software deployment.The time finally completed required for deployment is 48 minutes.Detecting the average transmission rate obtaining network in deployment is 990Mb.Image transmission time is 150 seconds, then according to calculating deployment time T=150+3700 × 8 of formula 1 × 87/990=2751 second=46 minutes.This time and the experiment statistics time substantially identical, prove that formula 1 is correct.
Can see from formula 1, the time required for deployment is linear with the destination node number installed simultaneously, and the coefficient of this linear relationship is determined by the size of videoing and the network bandwidth.Video larger, coefficient is larger, and the time required for deployment is longer; The network bandwidth is larger, and coefficient is less, and the time required for deployment is shorter.
Its processing and fabricating of the automated quick deployment method of large-scale computer cluster node of the present invention is very simple and convenient, can process to specifications.
Except the technical characteristic described in specification, be the known technology of those skilled in the art.
Claims (2)
1. the automated quick deployment method of large-scale computer cluster node, it is characterized in that based on client/server architecture, node software installs the mode adopted based on reflection, reflection is a complete system, containing the operating system needing to install on the target node and application software package; All reflections all store on the server; A management node is provided with as server in hardware structure needed for this method, destination node is connected with management node by commodity ethernet switch, destination node needs the network interface card and mainboard of supporting Pre-Boot eXecution Environment and WOL, destination node is connected with management node by commodity ethernet switch as the client computer in this structure;
1), management node is the server end in architecture, comprise the service end of server end for image transmission and Pre-Boot eXecution Environment agreement, be responsible for creating reflection, management reflection, store destination node information, WOL destination node and the various services required for deployment being provided;
2), destination node is client in architecture, obtains and disposes agency, dispose agency and be responsible for downloading from the management node as server video and being installed on the hard disk of this locality from management node;
3), the start-up course of destination node observes the Pre-Boot eXecution Environment agreement of Intel, management node provides dynamic host configuration protocol DHCP, simple file transfer protocol TFTP serve and service for transmitting reflection;
4), between destination node and management node, carry out image transmission in, use the different remote copy instruments based on client/server architecture;
The concrete steps of cluster node automated quick deployment method are:
Step 1: obtain reflection:
Reflection is obtained by following two kinds of modes:
The first creates reflection by installation kit, namely the installation kit of operating system in fixing disc and application software copied in a catalogue of management node, directly on management node, create virtual mapping with installation kit;
The second is from a source node snapping reflection, namely what a node of first manual installation, as the source node of reflection, source node is installed the various application software of operating system and needs, then catch the reflection of source node from management node and store, applicable destination node and source node are the situations of isomorphism;
Step 2: binding reflection and destination node:
The network identification information of destination node and reflection binds, determines to video and the corresponding relation of destination node, one of them corresponding multiple destination node of videoing, but destination node can only the unique reflection of correspondence;
Step 3: installation targets node:
When disposing, first on management node, start DHCP service, for destination node provides network identification information; Start TFTP service, download deployment Agent for destination node and service is provided; Be enabled in the server end of the remote copy instrument transmitting reflection between management node and destination node;
Management node starts destination node by Wakeup On LAN; After destination node is started by network interface card, first obtain network identity according to dynamic host configuration protocol DHCP from management node, then download according to simple file transfer protocol TFTP and dispose agency and run in local memory; Dispose agency and download reflection from management node, complete installation in this locality, after installation, node autoboot;
Step 4: configuration destination node:
When destination node installs initiating switchup, run configurator and the operating system of destination node and application software are configured; After configuration, namely destination node becomes a clustered node used immediately;
Step 5: test:
The all node installation of group system is complete, and testing results program, whether the Install and configure of Test Operating System and application software is correct.
2. the automated quick deployment method of large-scale computer cluster node according to claim 1, is characterized in that described destination node is placed through network interface card and starts in BIOS, its software installation process full automation; Destination node does not need the ancillary equipment such as keyboard, mouse, display, CD-ROM drive, floppy drive; All operations all complete on management node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410718742.8A CN104394223A (en) | 2014-12-03 | 2014-12-03 | Automatic rapid deployment method of large-scale computer cluster system nodes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410718742.8A CN104394223A (en) | 2014-12-03 | 2014-12-03 | Automatic rapid deployment method of large-scale computer cluster system nodes |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104394223A true CN104394223A (en) | 2015-03-04 |
Family
ID=52612058
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410718742.8A Pending CN104394223A (en) | 2014-12-03 | 2014-12-03 | Automatic rapid deployment method of large-scale computer cluster system nodes |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104394223A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105183520A (en) * | 2015-09-21 | 2015-12-23 | 赵伟 | Automatic remote installing and debugging method and system for computer software |
| CN105187252A (en) * | 2015-09-28 | 2015-12-23 | 浪潮(北京)电子信息产业有限公司 | Method and system for batch deployment of customized systems |
| CN105183512A (en) * | 2015-08-28 | 2015-12-23 | 浪潮(北京)电子信息产业有限公司 | Method and system for deploying operating system in physical hard disk |
| CN106293819A (en) * | 2016-08-03 | 2017-01-04 | 浪潮(北京)电子信息产业有限公司 | A kind of batch dispositions method based on snapshot and system |
| CN106533790A (en) * | 2016-12-12 | 2017-03-22 | 郑州云海信息技术有限公司 | Method, apparatus and system for deploying storage server, and storage server |
| CN106603316A (en) * | 2017-02-08 | 2017-04-26 | 北京启明星辰信息安全技术有限公司 | Cluster deploying method, first server, and second server |
| CN106681956A (en) * | 2016-12-27 | 2017-05-17 | 北京锐安科技有限公司 | Method and device for operating large-scale computer cluster |
| CN106686142A (en) * | 2017-03-07 | 2017-05-17 | 五道口科技（北京）股份有限公司 | Integrated Linux server system and system image push method thereof based on Web |
| CN106789168A (en) * | 2016-11-28 | 2017-05-31 | 华为技术有限公司 | A kind of data center server manages the dispositions method and frame top formula interchanger of network |
| CN107071043A (en) * | 2017-04-28 | 2017-08-18 | 曙光信息产业(北京)有限公司 | The collocation method and device of a kind of network information |
| CN109144523A (en) * | 2017-06-19 | 2019-01-04 | 中兴通讯股份有限公司 | Software deployment management method and server, management client, normal client end |
| CN109379223A (en) * | 2018-10-18 | 2019-02-22 | 郑州云海信息技术有限公司 | A kind of method and apparatus for realizing network interface card automated setting |
| CN112346649A (en) * | 2019-08-09 | 2021-02-09 | 北京国双科技有限公司 | Method and device for configuring disk |
| CN114500522A (en) * | 2020-10-23 | 2022-05-13 | 慧与发展有限责任合伙企业 | Shared image installation image stream |
-
2014
- 2014-12-03 CN CN201410718742.8A patent/CN104394223A/en active Pending
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105183512A (en) * | 2015-08-28 | 2015-12-23 | 浪潮(北京)电子信息产业有限公司 | Method and system for deploying operating system in physical hard disk |
| CN105183520A (en) * | 2015-09-21 | 2015-12-23 | 赵伟 | Automatic remote installing and debugging method and system for computer software |
| CN105183520B (en) * | 2015-09-21 | 2019-01-15 | 赵伟 | Computer software remote automation Method of Adjustment and system |
| CN105187252A (en) * | 2015-09-28 | 2015-12-23 | 浪潮(北京)电子信息产业有限公司 | Method and system for batch deployment of customized systems |
| CN106293819A (en) * | 2016-08-03 | 2017-01-04 | 浪潮(北京)电子信息产业有限公司 | A kind of batch dispositions method based on snapshot and system |
| CN106789168A (en) * | 2016-11-28 | 2017-05-31 | 华为技术有限公司 | A kind of data center server manages the dispositions method and frame top formula interchanger of network |
| CN106789168B (en) * | 2016-11-28 | 2020-03-20 | 华为技术有限公司 | Deployment method of data center server management network and rack top type switch |
| CN106533790A (en) * | 2016-12-12 | 2017-03-22 | 郑州云海信息技术有限公司 | Method, apparatus and system for deploying storage server, and storage server |
| CN106681956A (en) * | 2016-12-27 | 2017-05-17 | 北京锐安科技有限公司 | Method and device for operating large-scale computer cluster |
| CN106603316B (en) * | 2017-02-08 | 2019-05-14 | 北京启明星辰信息安全技术有限公司 | A kind of dispositions method of cluster, first server and second server |
| CN106603316A (en) * | 2017-02-08 | 2017-04-26 | 北京启明星辰信息安全技术有限公司 | Cluster deploying method, first server, and second server |
| CN106686142A (en) * | 2017-03-07 | 2017-05-17 | 五道口科技（北京）股份有限公司 | Integrated Linux server system and system image push method thereof based on Web |
| CN107071043A (en) * | 2017-04-28 | 2017-08-18 | 曙光信息产业(北京)有限公司 | The collocation method and device of a kind of network information |
| CN109144523A (en) * | 2017-06-19 | 2019-01-04 | 中兴通讯股份有限公司 | Software deployment management method and server, management client, normal client end |
| CN109379223A (en) * | 2018-10-18 | 2019-02-22 | 郑州云海信息技术有限公司 | A kind of method and apparatus for realizing network interface card automated setting |
| CN112346649A (en) * | 2019-08-09 | 2021-02-09 | 北京国双科技有限公司 | Method and device for configuring disk |
| CN114500522A (en) * | 2020-10-23 | 2022-05-13 | 慧与发展有限责任合伙企业 | Shared image installation image stream |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104394223A (en) | Automatic rapid deployment method of large-scale computer cluster system nodes | |
| US9081747B1 (en) | Computer program deployment to one or more target devices | |
| CN1731738A (en) | Automated quick deployment method for large-scale computer cluster node | |
| CN101840346B (en) | Method and system for deploying cloud host computer | |
| US7600005B2 (en) | Method and apparatus for provisioning heterogeneous operating systems onto heterogeneous hardware systems | |
| RU2417415C2 (en) | Method and system for deploying software, software deployment server and user server | |
| TWI592808B (en) | High-speed automated cluster system deployment using virtual disks | |
| US8943183B2 (en) | Decoupled installation of data management systems | |
| US8332496B2 (en) | Provisioning of operating environments on a server in a networked environment | |
| US8352916B2 (en) | Facilitating the automated testing of daily builds of software | |
| CN1848787A (en) | Automatic fast dispositioning method for aggregated server system node | |
| CN101820387B (en) | Method for rapidly deploying extensible cluster | |
| CN101393524A (en) | Firmware update method and system using the same | |
| CN106911729A (en) | A kind of operating system remote installation method suitable for domestic processor | |
| US10372463B1 (en) | Provisioning a computerized device with an operating system | |
| CN111198696B (en) | OpenStack large-scale deployment method and system based on bare computer server | |
| JP2009070375A (en) | System and method for custom installation of operating system on remote client | |
| CN105955742A (en) | BIOS refreshing method, service system and BIOS refreshing system | |
| CN102591679A (en) | Quick application software arrangement method for cluster system | |
| CN105183520B (en) | Computer software remote automation Method of Adjustment and system | |
| CN108984272A (en) | The control method and device of OpenStack cloud computing management platform | |
| CN109254783A (en) | A kind of server system installation method, device, equipment and readable storage medium storing program for executing | |
| US7882232B2 (en) | Rapid resource provisioning with automated throttling | |
| CN105119732A (en) | Device and method for distributing hosts in cluster environment | |
| US11528186B2 (en) | Automated initialization of bare metal servers |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150304 |
|
| WD01 | Invention patent application deemed withdrawn after publication |