CN102594602A - Location management design method of multi-node cloud computing server - Google Patents
Location management design method of multi-node cloud computing server Download PDFInfo
- Publication number
- CN102594602A CN102594602A CN201210041439XA CN201210041439A CN102594602A CN 102594602 A CN102594602 A CN 102594602A CN 201210041439X A CN201210041439X A CN 201210041439XA CN 201210041439 A CN201210041439 A CN 201210041439A CN 102594602 A CN102594602 A CN 102594602A
- Authority
- CN
- China
- Prior art keywords
- node
- slot
- server
- module
- plate
- 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.)
- Granted
Links
Images
Landscapes
- Small-Scale Networks (AREA)
- Power Sources (AREA)
Abstract
The invention provides a location management design method of a multi-node cloud computing server, wherein the hardware comprises a server cabinet, a computing node slot, a 3+1 power module slot, a Gigabit switch extension reserving slot, a cable guide slot, a 3+1 power module slot, a Gigabit switch data network module slot, a RMC (rod memory computer) module slot, a Gigabit switch management network slot, a reserving power module function extension slot, a reserving power module function extension slot, a hot-plug fan module, middle node plates and a power strip, wherein in the server cabinet, 80 server nodes can be supported in a maximum manner, which can be divided into four layers from top to bottom; one middle node plate exists in each layer; power supplying and signal interconnection of 20 nodes in the layer can be achieved; each server node is connected with the middle node plate by a secondary power plate; an I2C link is reversed on each node and is connected with 20 golden finger connectors on the middle node plates by the secondary power plate; at least 20 server nodes can be supported by one middle node plate so that interconnection and power supplying of the nodes are achieved.
Description
Technical field
The present invention relates to a kind of computer server technical field, specifically a kind of multinode cloud computing server orientation management method for designing.
Summary of the invention
The purpose of this invention is to provide a kind of multinode cloud computing server orientation management method for designing.
The objective of the invention is to realize by following mode; Comprise server cabinet, computing node slot, 3+1 power module slot, gigabit switch expansion reservation slot, netting twine gathering sill, 3+1 power module slot, gigabit switched data network module slot, RMC module slot, gigabit exchange of management net slot, reserve power module function expansion slot, reserve plate, electric power plug socket in power module function expansion slot, hot-swappable fan module, the node; In server cabinet; 80 server nodes of maximum support; Being divided into from top to bottom is four layers, and every layer has plate in a slice node, realizes that the power supply and the signal of 20 node of this layer is interconnected; Each server node is interconnected through plate in secondary power supply plate and the node, all is reserved with the I2C link on each node, is connected in the node on 20 golden finger connectors on the plate through the secondary power supply plate; Plate is supported 20 server nodes at least in a slice node, is used to realize the interconnected of node and power supply;
Design procedure is following:
1) server cabinet four layers of design from the top down, each layer top is provided with the netting twine gathering sill, perpendicular slotting ten the computing node slots in left side, the 1st layer of right side is provided with a power module and reserves slot and a gigabit switch expansion reservation slot; The 2nd layer of right side is provided with a 3+1 power module slot and a gigabit switched data network module slot; Between the 3rd layer and the second layer two horizontal RMC module slot slots are set; The 4th layer of right side is provided with a 3+1 power module slot slot and a gigabit exchange of management net module slot; The 4th layer is that the basecoat right side is provided with a reservation gigabit switching network expansion slot and a reservation power module function expansion slot; The server cabinet rear portion is provided with four hot-swappable fan modules; Corresponding with the four-level space of server cabinet respectively; Between hot-swappable fan module and server cabinet, be horizontally arranged with plate in the node; The equipment that plate is provided with in data line interface and the server cabinet in the node is connected, and the server cabinet side also is provided with electric power plug socket enforcement centrally connected power supply; Through the centrally connected power supply of above-mentioned design implementation rack, concentrated heat radiation and centralized management, and then the work of simplification O&M is put, is disposed fast in the unification of realization multinode on the shelf;
2) each server node inside is combined into node unit the most basic in the multi node server through server master board, hard disk, internal memory, CPU basic module:
Based on I2C communication protocol; The GPIO signal of 20 golden finger connectors in the node on the plate through drawing on the RMC module; Realize unique address addressing; No matter server node is found on that golden finger connector like this, and the corresponding with it positioning function that realizes server node of unique address addressing is all arranged;
Based on I2C communication protocol; RMC module in the server cabinet is through the address strobe function of GPIO; Interconnected with the I2C link of 80 nodes; By means of IPMB standard agreement packet, the BMC of RMC module and each node carries out real time communication, obtains the assets information of each computing node, grasps the work state information of each sensor information, internal memory and the CPU of all nodes in real time.
The invention has the beneficial effects as follows: industry writes down the assets information and the placement location of server often through manual type for the server of newly putting on the shelf at present; When server failure; Need the artificial server apparatus placement location of searching corresponding assets information, bring and dispose the reduction of putting on the shelf, adopt this method with the O&M operating efficiency; Can simplify this type of work greatly, increase work efficiency.The application model of server apparatus is towards centralization, large-scale development, and how the location and the asset management of Rapid Realization server become following primary study content in the server admin field.
Description of drawings
Fig. 1 is the forward sight structural representation of rack;
Fig. 2 is the backsight structural representation of rack;
Fig. 3 is the backsight structural representation that rack removes fan board;
Fig. 4 is the server theory diagram.
Description of reference numerals: computing node slot 1,3+1 power module slot 2, gigabit switch expansion are reserved slot 3, netting twine gathering sill 4,3+1 power module slot 5, gigabit switched data network module slot 6, RMC module slot 7, gigabit exchange of management net slot 8, reserve power module function expansion slot 9, are reserved plate 12, electric power plug socket 13, server cabinet 14, data line interface 15 in power module function expansion slot 10, hot-swappable fan module 11, the node.
Embodiment
Explanation at length below with reference to Figure of description method of the present invention being done.
Help weapon hardware to form by plate 12, electric power plug socket 13 in server cabinet 14, computing node slot 1,3+1 power module slot 2, gigabit switch expansion reservation slot 3, netting twine gathering sill 4,3+1 power module slot 5, gigabit switched data network module slot 6, RMC module slot 7, gigabit exchange of management net slot 8, reservation power module function expansion slot 9, reservation power module function expansion slot 10, hot-swappable fan module 11, the node
1) server cabinet 14 four layers of design from the top down, each layer top is provided with netting twine gathering sill 4, and perpendicular 1, the 1 layer of right side of ten computing node slots of inserting, left side is provided with a power module reservation slot 2 and a gigabit switch expansion reservation slot 3; The 2nd layer of right side is provided with a 3+1 power module slot 5 and a gigabit switched data network module slot 6 respectively; Between the 3rd layer and the second layer two horizontal RMC module slot slots 7 are set; The 4th layer of right side is provided with a 3+1 power module slot slot 5 and a gigabit exchange of management net module slot 8; The 4th layer is that the basecoat right side is provided with a reservation gigabit switching network expansion slot 9 and a reservation power module function expansion slot 10; Server cabinet 14 rear portions are provided with four hot-swappable fan modules 11; Corresponding with the four-level space of server cabinet 14 respectively; Between hot-swappable fan module 11 and server cabinet 14, be horizontally arranged with plate 12 in the node; The equipment that plate 12 is provided with in data line interface 15 and the server cabinet 14 in the node are connected, and server cabinet 14 sides also are provided with electric power plug socket 13 enforcement centrally connected power supplies.
Embodiment
In Smart Rack, 80 server nodes of maximum support, being divided into from top to bottom is four layers, every layer has plate in a slice node, realizes that the power supply and the signal of 20 node of this layer is interconnected.
Shown in accompanying drawing, in server system, 80 server nodes of maximum support, being divided into from top to bottom is four layers, every layer has plate in a slice node, realizes that the power supply and the signal of 20 node of this layer is interconnected.
Each server node is interconnected through plate in the firm and hard existing and node of secondary power supply.All be reserved with the I2C link on each node, be connected to through the secondary power supply plate on the golden finger connector of plate in the node; And the plate maximum can be supported 20 server nodes in a slice node, and promptly the top has 20 golden finger connectors, is used to realize the interconnected of node and power supply.
The GPIO signal of 20 golden finger connectors in the node on the plate through drawing on the RMC; Realize unique address addressing; No matter server node is found on that golden finger connector like this, all has unique address addressing corresponding with it, so realizes the positioning function of server node;
RMC in the multi node server is through the address strobe function of GPIO; Interconnected with the I2C link of 80 nodes; By means of IPMB standard agreement packet; RMC can carry out real time communication with the BMC of each node, not only can obtain the assets information of each node, can also grasp each sensor information and the work state informations such as internal memory, CPU of all nodes in real time.
The present invention mainly is a kind of based on the automatic location of I2C communication protocol realization multiserver node and the method for asset management.
1) through the centralized management terminal, for example the RMC administration module of our exploitation through the GPIO signal, is to encode in the location groove position of 80 nodes, realizes the location of server;
2) through the I2C communication link, the gating function of collocation GPIO realizes the Card read/write of RMC to 80 nodes being attached thereto, thereby obtains the automatic summarizing and reporting of assets information of Servers-all node;
Except that the described technical characterictic of specification, be the known technology of those skilled in the art.
Claims (1)
1. multinode cloud computing server orientation management method for designing; It is characterized in that hardware comprises server cabinet, computing node slot, 3+1 power module slot, gigabit switch expansion reservation slot, netting twine gathering sill, 3+1 power module slot, gigabit switched data network module slot, RMC module slot, gigabit exchange of management net slot, reserves power module function expansion slot, reserves plate, electric power plug socket in power module function expansion slot, hot-swappable fan module, the node; In server cabinet; 80 server nodes of maximum support; Be divided into from top to bottom is four layers; Every layer has plate in a slice node, realizes that the power supply and the signal of 20 node of this layer is interconnected; Each server node is interconnected through plate in secondary power supply plate and the node, all is reserved with the I2C link on each node, is connected in the node on 20 golden finger connectors on the plate through the secondary power supply plate; Plate is supported 20 server nodes at least in a slice node, is used to realize the interconnected of node and power supply; Design procedure is following:
1) server cabinet four layers of design from the top down, each layer top is provided with the netting twine gathering sill, perpendicular slotting ten the computing node slots in left side, the 1st layer of right side is provided with a power module and reserves slot and a gigabit switch expansion reservation slot; The 2nd layer of right side is provided with a 3+1 power module slot and a gigabit switched data network module slot; Between the 3rd layer and the second layer two horizontal RMC module slot slots are set; The 4th layer of right side is provided with a 3+1 power module slot slot and a gigabit exchange of management net module slot; The 4th layer is that the basecoat right side is provided with a reservation gigabit switching network expansion slot and a reservation power module function expansion slot; The server cabinet rear portion is provided with four hot-swappable fan modules; Corresponding with the four-level space of server cabinet respectively; Between hot-swappable fan module and server cabinet, be horizontally arranged with plate in the node; The equipment that plate is provided with in data line interface and the server cabinet in the node is connected, and the server cabinet side also is provided with electric power plug socket enforcement centrally connected power supply; Through the centrally connected power supply of above-mentioned design implementation rack, concentrated heat radiation and centralized management, and then the work of simplification O&M is put, is disposed fast in the unification of realization multinode on the shelf;
2) each server node inside is combined into node unit the most basic in the multi node server through server master board, hard disk, internal memory, CPU basic module:
Based on I2C communication protocol; The GPIO signal of 20 golden finger connectors in the node on the plate through drawing on the RMC module; Realize unique address addressing; No matter server node is found on that golden finger connector like this, and the corresponding with it positioning function that realizes server node of unique address addressing is all arranged;
Based on I2C communication protocol; RMC module in the server cabinet is through the address strobe function of GPIO; Interconnected with the I2C link of 80 nodes; By means of IPMB standard agreement packet, the BMC of RMC module and each node carries out real time communication, obtains the assets information of each computing node, grasps the work state information of each sensor information, internal memory and the CPU of all nodes in real time.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210041439.XA CN102594602B (en) | 2012-02-23 | 2012-02-23 | A kind of location management design method of multi-node cloud computing server |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210041439.XA CN102594602B (en) | 2012-02-23 | 2012-02-23 | A kind of location management design method of multi-node cloud computing server |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102594602A true CN102594602A (en) | 2012-07-18 |
CN102594602B CN102594602B (en) | 2016-08-10 |
Family
ID=46482822
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210041439.XA Active CN102594602B (en) | 2012-02-23 | 2012-02-23 | A kind of location management design method of multi-node cloud computing server |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102594602B (en) |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102902322A (en) * | 2012-09-14 | 2013-01-30 | 珠海报业文化传播有限公司 | Cabinet device of cloud computing system |
CN103118103A (en) * | 2013-01-29 | 2013-05-22 | 浪潮电子信息产业股份有限公司 | Cloud server framework capable of achieving multi-node interconnection and management |
CN103532762A (en) * | 2013-10-22 | 2014-01-22 | 浪潮电子信息产业股份有限公司 | I2C link switching-based server asset information management design method |
CN103983876A (en) * | 2014-05-27 | 2014-08-13 | 浪潮电子信息产业股份有限公司 | Method for online detection of node middle panel |
CN104407680A (en) * | 2014-12-02 | 2015-03-11 | 浪潮电子信息产业股份有限公司 | Management system based on complete cabinet nodes |
CN104597977A (en) * | 2015-01-13 | 2015-05-06 | 浪潮电子信息产业股份有限公司 | Front maintenance type 2U4 node high-density server system |
TWI497312B (en) * | 2013-12-19 | 2015-08-21 | Inventec Corp | High density server system |
CN105446199A (en) * | 2015-12-31 | 2016-03-30 | 北京华胜天成科技股份有限公司 | Intelligent cabinet |
CN105490859A (en) * | 2015-12-18 | 2016-04-13 | 山东海量信息技术研究院 | Node positioning method of high-end fault tolerance server |
CN106294230A (en) * | 2016-08-10 | 2017-01-04 | 浪潮电子信息产业股份有限公司 | Method for expanding communication secondary command between RMC and middle plate |
WO2017063331A1 (en) * | 2015-10-14 | 2017-04-20 | 浪潮电子信息产业股份有限公司 | Fan speed-adjustment policy for entire machine cabinet by placing fan table on node bmc |
CN106850814A (en) * | 2017-02-15 | 2017-06-13 | 济南浪潮高新科技投资发展有限公司 | It is a kind of to increase the method that custom command is supported to realize sensor information collection |
CN106919233A (en) * | 2017-02-25 | 2017-07-04 | 郑州云海信息技术有限公司 | A kind of high density storage server architecture system |
CN107241886A (en) * | 2017-06-06 | 2017-10-10 | 郑州云海信息技术有限公司 | Plate and its design method in a kind of new Rack nodes |
CN109213285A (en) * | 2018-09-27 | 2019-01-15 | 郑州云海信息技术有限公司 | A kind of edge calculations server architecture |
CN111209241A (en) * | 2019-12-19 | 2020-05-29 | 曙光信息系统(辽宁)有限公司 | Management system of whole cabinet server |
CN111315176A (en) * | 2020-02-29 | 2020-06-19 | 苏州浪潮智能科技有限公司 | Method and device for positioning nodes in cabinet |
CN113645067A (en) * | 2021-08-03 | 2021-11-12 | 加弘科技咨询(上海)有限公司 | Network management system and method of multi-node substrate management controller |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030126253A1 (en) * | 1996-07-23 | 2003-07-03 | Ewing Carrel W. | Network remote power management outlet strip |
CN101030880A (en) * | 2006-03-03 | 2007-09-05 | 环达电脑(上海)有限公司 | Multi-server management system and method |
CN201698312U (en) * | 2010-07-26 | 2011-01-05 | 深圳市腾讯计算机系统有限公司 | Cooling system for computer case |
-
2012
- 2012-02-23 CN CN201210041439.XA patent/CN102594602B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030126253A1 (en) * | 1996-07-23 | 2003-07-03 | Ewing Carrel W. | Network remote power management outlet strip |
CN101030880A (en) * | 2006-03-03 | 2007-09-05 | 环达电脑(上海)有限公司 | Multi-server management system and method |
CN201698312U (en) * | 2010-07-26 | 2011-01-05 | 深圳市腾讯计算机系统有限公司 | Cooling system for computer case |
Cited By (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102902322A (en) * | 2012-09-14 | 2013-01-30 | 珠海报业文化传播有限公司 | Cabinet device of cloud computing system |
CN103118103A (en) * | 2013-01-29 | 2013-05-22 | 浪潮电子信息产业股份有限公司 | Cloud server framework capable of achieving multi-node interconnection and management |
CN103532762A (en) * | 2013-10-22 | 2014-01-22 | 浪潮电子信息产业股份有限公司 | I2C link switching-based server asset information management design method |
TWI497312B (en) * | 2013-12-19 | 2015-08-21 | Inventec Corp | High density server system |
CN103983876B (en) * | 2014-05-27 | 2017-12-01 | 浪潮电子信息产业股份有限公司 | A kind of method of plate on-line checking in node |
CN103983876A (en) * | 2014-05-27 | 2014-08-13 | 浪潮电子信息产业股份有限公司 | Method for online detection of node middle panel |
CN104407680A (en) * | 2014-12-02 | 2015-03-11 | 浪潮电子信息产业股份有限公司 | Management system based on complete cabinet nodes |
CN104597977A (en) * | 2015-01-13 | 2015-05-06 | 浪潮电子信息产业股份有限公司 | Front maintenance type 2U4 node high-density server system |
US10405455B2 (en) | 2015-10-14 | 2019-09-03 | Inspur Electronic Information Industry Co., Ltd. | Fan speed-adjustment policy for entire machine cabinet by placing fan table on node BMC |
WO2017063331A1 (en) * | 2015-10-14 | 2017-04-20 | 浪潮电子信息产业股份有限公司 | Fan speed-adjustment policy for entire machine cabinet by placing fan table on node bmc |
CN105490859A (en) * | 2015-12-18 | 2016-04-13 | 山东海量信息技术研究院 | Node positioning method of high-end fault tolerance server |
CN105490859B (en) * | 2015-12-18 | 2019-01-22 | 山东海量信息技术研究院 | A kind of node positioning method of high-end fault-tolerant server |
CN105446199A (en) * | 2015-12-31 | 2016-03-30 | 北京华胜天成科技股份有限公司 | Intelligent cabinet |
CN105446199B (en) * | 2015-12-31 | 2018-11-16 | 北京华胜天成科技股份有限公司 | A kind of intelligent cabinet |
CN106294230A (en) * | 2016-08-10 | 2017-01-04 | 浪潮电子信息产业股份有限公司 | Method for expanding communication secondary command between RMC and middle plate |
CN106850814A (en) * | 2017-02-15 | 2017-06-13 | 济南浪潮高新科技投资发展有限公司 | It is a kind of to increase the method that custom command is supported to realize sensor information collection |
CN106850814B (en) * | 2017-02-15 | 2020-02-14 | 浪潮集团有限公司 | Method for realizing sensor information acquisition by adding custom command support |
CN106919233A (en) * | 2017-02-25 | 2017-07-04 | 郑州云海信息技术有限公司 | A kind of high density storage server architecture system |
CN107241886A (en) * | 2017-06-06 | 2017-10-10 | 郑州云海信息技术有限公司 | Plate and its design method in a kind of new Rack nodes |
CN107241886B (en) * | 2017-06-06 | 2020-02-11 | 苏州浪潮智能科技有限公司 | Novel Rack node middle plate and design method thereof |
CN109213285A (en) * | 2018-09-27 | 2019-01-15 | 郑州云海信息技术有限公司 | A kind of edge calculations server architecture |
CN111209241A (en) * | 2019-12-19 | 2020-05-29 | 曙光信息系统(辽宁)有限公司 | Management system of whole cabinet server |
CN111315176A (en) * | 2020-02-29 | 2020-06-19 | 苏州浪潮智能科技有限公司 | Method and device for positioning nodes in cabinet |
CN111315176B (en) * | 2020-02-29 | 2021-07-06 | 苏州浪潮智能科技有限公司 | Method and device for positioning nodes in cabinet |
CN113645067A (en) * | 2021-08-03 | 2021-11-12 | 加弘科技咨询(上海)有限公司 | Network management system and method of multi-node substrate management controller |
Also Published As
Publication number | Publication date |
---|---|
CN102594602B (en) | 2016-08-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102594602A (en) | Location management design method of multi-node cloud computing server | |
KR102542562B1 (en) | Modular non-volatile flash memory blade | |
CN108430759A (en) | The rack-mounted storage assembly of grid type framework | |
US10372360B2 (en) | Apparatus, system, and method for reconfigurable media-agnostic storage | |
JP3157935U (en) | server | |
US10420246B2 (en) | Modular computer system and server module | |
US10499530B2 (en) | Server system | |
CN104049698A (en) | Storage and compute server based on 6U space | |
CN102625608A (en) | Design method for large-scale multiple-node server cabinets | |
CN102567227A (en) | Double-controller memory system and method for sharing cache equipment | |
CN104967577B (en) | SAS switch and server | |
CN104407680A (en) | Management system based on complete cabinet nodes | |
CN104639469A (en) | Computing and storing cluster system based on PCIE (Peripheral Component Interconnect Express) interconnection | |
CN104951022A (en) | Backplane connecting system and method for blade server | |
CN211062032U (en) | Scorpio high-performance storage node based on domestic Shenwei platform | |
CN106997224A (en) | Hard disk subrack and server | |
CN102799708B (en) | Graphic processing unit (GPU) high-performance calculation platform device applied to electromagnetic simulation | |
CN105577752B (en) | A kind of management system for fusion architecture server | |
CN104156037A (en) | High-density micro server based on cloud services | |
CN103746941A (en) | Board-level interconnected big data all-in-one machine | |
CN204117069U (en) | A kind of server backplanes | |
CN203025686U (en) | Multi-source data storage device | |
WO2016086700A1 (en) | Rack and communication method | |
CN107491408B (en) | Computing server node | |
CN104932837A (en) | Storage pool framework |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |