CN106528488B - Computing cluster system and control method in a kind of single machine - Google Patents
Computing cluster system and control method in a kind of single machine Download PDFInfo
- Publication number
- CN106528488B CN106528488B CN201610883006.7A CN201610883006A CN106528488B CN 106528488 B CN106528488 B CN 106528488B CN 201610883006 A CN201610883006 A CN 201610883006A CN 106528488 B CN106528488 B CN 106528488B
- Authority
- CN
- China
- Prior art keywords
- cpu
- distribution
- services module
- network interface
- interface card
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F15/00—Digital computers in general; Data processing equipment in general
- G06F15/16—Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
- G06F15/161—Computing infrastructure, e.g. computer clusters, blade chassis or hardware partitioning
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Multi Processors (AREA)
- Computer And Data Communications (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses computing cluster system and control methods in a kind of single machine, the system includes distribution of services module, muti-piece network interface card, CPU and multiple hard disks, the CPU includes multiple calculating cores, multiple calculating cores of the muti-piece network interface card and CPU correspond multiple calculate nodes in composition single machine, and the CPU is connect with multiple hard disks by data/address bus.The present invention by single machine by multiple calculating cores, multiple Network card setups be multiple calculate nodes, and it is connected and bound with multiple hard disks, the reasonable distribution of application request is realized by distribution of services module, it is realized with a low cost high IO performance and efficient PC cluster in single machine, and perfect can be adapted to multiple access nets.The present invention can be widely applied to data processing field as computing cluster system in a kind of single machine and control method.
Description
Technical field
The present invention relates to data processing field, computing cluster system and control method in especially a kind of single machine.
Background technique
Current all high-performance, high IO computing cluster are all based on more physical servers, a server conduct
One calculate node, to form computing cluster.Due to relying on more physical servers, cost is relatively high.
To solve the above-mentioned problems, in the prior art in a kind of cluster single machine multinode implementation method
(CN102999571B) it by disposing multiple nodes on a server, no longer needs to rely on more physical servers, but its side
Method only illustrates the multiple nodes and server disk of deployment, is accomplished that the shared of all nodal informations in cluster, solves
It is unable to fully the limitation using muti-piece disk, but and increases processing capacity；Especially in the application of mass data IO,
Data volume is big due to read-write, miscellaneous more than type, and server is frequently cut between a variety of different data, between read-write operation
It changes, leads to that treatment effeciency is low, error rate is high.
Summary of the invention
In order to solve the above-mentioned technical problem, the object of the present invention is to provide high-performance, high IO speed are realized in a kind of single machine
The computing cluster system of degree.
In order to solve the above-mentioned technical problem, it is another object of the present invention to: provide and realize high-performance, height in a kind of single machine
The computing cluster control method of IO speed.
The technical scheme adopted by the invention is that: computing cluster system in a kind of single machine, include distribution of services module,
Muti-piece network interface card, CPU and multiple hard disks, the CPU include multiple calculating cores, the muti-piece network interface card and CPU's
Multiple cores that calculate correspond the multiple calculate nodes formed in single machine, and the CPU and multiple hard disks pass through data/address bus and connect
It connects.
Further, the distribution of services module is DNS load balancing module.
Further, the distribution of services module is HTTP scheduler module.
Further, the multiple hard disk is SSD hard disk.
Further, the CPU more calculate each calculating core and at least one piece of hard disk binding in core.
It is of the present invention another solution is that computing cluster control method in a kind of single machine, includes following step
It is rapid:
A, the CPU in single machine is calculated to core more connects the multiple calculate nodes of composition, each calculate node with muti-piece network interface card
It include that a CPU calculating core and one piece of network interface card, each calculate node at least bind one piece of hard disk；
B, distribution of services module obtains application request, and request content is sent to corresponding calculate node；
C, the network interface card in calculate node receives request content and is forwarded to corresponding CPU calculating core and handled, then
It is written and read by hard disk of the data/address bus to binding.
Further, the CPU of each calculate node calculates core and is provided with origin server mark in the step A, described
Origin server is identified for distinguishing application request source.
Further, the step B specifically: distribution of services module obtains application request, and by request content through overload
Corresponding calculate node is sent to after balanced or HTTP scheduling.
Further, the step B specifically includes following sub-step:
B1, distribution of services module obtain application request；
Application request source is sent to CPU by the network interface card of each calculate node and calculates core by B2, distribution of services module；
B3, CPU calculate core and indicate to compare by application request source and origin server, and comparing result is identical CPU
It calculates core and passes through network interface card feedback result to distribution of services module；
B4, distribution of services module are fed back according to above-mentioned steps B3 is sent to corresponding net for the request content of application request
Card.
Further, the CPU of each calculate node calculates core and is additionally provided with dispatching priority mark, institute in the step A
State step B4 specifically:
If having multiple CPU to calculate core in step B3 passes through network interface card feedback result to distribution of services module, distribution of services
Module identifies according to above-mentioned dispatching priority the request content of application request being sent to corresponding network interface card.
The beneficial effects of the present invention are: multiple calculating cores, multiple Network card setups are more in single machine by present system
A calculate node, and connected and bound with multiple hard disks, the reasonable distribution of application request, low cost are realized by distribution of services module
It realizes high IO performance and efficient PC cluster in single machine, and perfect can be adapted to multiple access nets.
Another beneficial effect of the present invention is: the method for the present invention is in single machine by multiple calculating cores, multiple Network card setups
It for multiple calculate nodes, and is connected and bound with multiple hard disks, the reasonable distribution of application request is realized by distribution of services module, it is low
Cost realizes high IO performance and efficient PC cluster in single machine, and perfect can be adapted to multiple access nets.
Detailed description of the invention
Fig. 1 is system structure schematic block diagram in single machine in the present invention；
Fig. 2 is cluster control method flow chart of steps in single machine in the present invention.
Specific embodiment
Specific embodiments of the present invention will be further explained with reference to the accompanying drawing:
Referring to Fig.1, computing cluster system in a kind of single machine, include distribution of services module,
Muti-piece network interface card, CPU and multiple hard disks, the CPU include multiple calculating cores, the muti-piece network interface card and CPU's
Multiple cores that calculate correspond the multiple calculate nodes formed in single machine, and the CPU and multiple hard disks pass through data/address bus and connect
It connects.
The CPU can be used: 1, multiple processors, and each processor is as a calculating core；2, multi-core processor is made
For multiple calculating cores；3, other more calculating core technologies.
It is further used as preferred embodiment, the distribution of services module is DNS load balancing module.
It is further used as preferred embodiment, the distribution of services module is HTTP scheduler module.
It is further used as preferred embodiment, the multiple hard disk is SSD hard disk.
It is further used as preferred embodiment, the CPU more calculate in cores each calculatings core and at least one piece
Hard disk binding.
Referring to Fig. 2, computing cluster control method in a kind of single machine, wherein computing cluster system structure can refer to figure in single machine
1, include following steps:
A, the CPU in single machine is calculated to core more connects the multiple calculate nodes of composition, each calculate node with muti-piece network interface card
It include that a CPU calculating core and one piece of network interface card, each calculate node at least bind one piece of hard disk；
CPU calculating core amounts wherein in single machine are expansible, and network interface card quantity is increase accordingly, and hard disk quantity can also be correspondingly
Increase.
B, distribution of services module obtains application request, and request content is sent to corresponding calculate node；
C, the network interface card in calculate node receives request content and is forwarded to corresponding CPU calculating core and handled, then
It is written and read by hard disk of the data/address bus to binding；
The CPU, which calculates core, can bind one or more hard disk, but one piece of hard disk only binds unique calculating core；Meter
Calculating can be bound between core and hard disk by simple mapping relations, when the calculating core bound in the hard disk is not utilized,
Binding relationship can be adjusted flexibly, I/O operation amount is more, frequent higher calculating core to be assigned to.
It is further used as preferred embodiment, the CPU of each calculate node calculates core and be provided in the step A
Source server mark, the origin server mark is for distinguishing application request source.
Each node in single machine only handles corresponding data source, calculates core appointing in calculating process to reduce
Business handoff procedure, improves processing capacity and I/O rate.
It is further used as preferred embodiment, the step B specifically: distribution of services module obtains application request, and
Request content is sent to corresponding calculate node after load balancing or HTTP scheduling.
For example, by using DNS load-balancing technique, the two-wire that computing cluster system deployment as shown in Figure 1 arrives is most common double
In line computer room, calculate node A(includes network interface card A and calculates core A) connect by the network interface card A of oneself network of telecommunications, calculating
Node B accordingly by network interface card B connection connection network, by the scheduling of DNS, so that the user of telecommunications accesses calculate node
User's accessed node of A, connection calculate B, and whole computer can be made to reach the maximum utilization of resources, including network I/O, magnetic
Disk IO and CPU etc..
It is further used as preferred embodiment, the step B specifically includes following sub-step:
B1, distribution of services module obtain application request；
Application request source is sent to CPU by the network interface card of each calculate node and calculates core by B2, distribution of services module；
B3, CPU calculate core and indicate to compare by application request source and origin server, and comparing result is identical CPU
It calculates core and passes through network interface card feedback result to distribution of services module；
B4, distribution of services module are fed back according to above-mentioned steps B3 is sent to corresponding net for the request content of application request
Card.
It is further used as preferred embodiment, the CPU of each calculate node calculates core and is additionally provided in the step A
Dispatching priority mark, the step B4 specifically:
If having multiple CPU to calculate core in step B3 passes through network interface card feedback result to distribution of services module, distribution of services
Module identifies according to above-mentioned dispatching priority the request content of application request being sent to corresponding network interface card.The dispatching priority
It can be set according to access frequency.
It is to be illustrated to preferable implementation of the invention, but the invention is not limited to the implementation above
Example, those skilled in the art can also make various equivalents on the premise of without prejudice to spirit of the invention or replace
It changes, these equivalent deformations or replacement are all included in the scope defined by the claims of the present application.
Claims (8)
1. computing cluster system in a kind of single machine, it is characterised in that: include distribution of services module, muti-piece network interface card, CPU and multiple
Hard disk, the CPU include multiple calculating cores, multiple calculating cores one-to-one correspondence composition list of the muti-piece network interface card and CPU
Multiple calculate nodes in machine, the CPU are connect with multiple hard disks by data/address bus；
Each calculate node at least binds one piece of hard disk；
Request content is sent to corresponding calculate node for obtaining application request by the distribution of services module；
The network interface card is used for the request content received and is forwarded to corresponding CPU calculating core；
The CPU calculates core for handling the request content received, and by data/address bus to the hard disk of binding
It is written and read；
Each CPU calculates core and is provided with origin server mark, and the origin server mark is for distinguishing application request
Source；
Request content is sent to corresponding calculate node especially by following steps by the distribution of services module:
Application request source is sent to CPU by the network interface card of each calculate node and calculates core by distribution of services module；
CPU calculates core and indicates to compare by application request source and origin server, and comparing result is that identical CPU calculates core
Pass through network interface card feedback result to distribution of services module；
The request content of application request is sent to corresponding network interface card according to the feedback that CPU calculates core by distribution of services module.
2. computing cluster system in a kind of single machine according to claim 1, it is characterised in that: the distribution of services module is
DNS load balancing module.
3. computing cluster system in a kind of single machine according to claim 1, it is characterised in that: the distribution of services module is
HTTP scheduler module.
4. computing cluster system in a kind of single machine according to claim 1, it is characterised in that: the multiple hard disk is SSD
Hard disk.
5. computing cluster system in a kind of single machine according to claim 1, it is characterised in that: the CPU more calculate core
Each calculating core is bound at least one piece of hard disk in the heart.
6. computing cluster control method in a kind of single machine, which is characterized in that include following steps:
A, the CPU in single machine is calculated to core more connects the multiple calculate nodes of composition with muti-piece network interface card, each calculate node includes
There are a CPU calculating core and one piece of network interface card, each calculate node at least to bind one piece of hard disk；
B, distribution of services module obtains application request, and request content is sent to corresponding calculate node；
C, the network interface card in calculate node receives request content and is forwarded to corresponding CPU calculating core and handled, and then passes through
Data/address bus is written and read the hard disk of binding；
The CPU of each calculate node calculates core and is provided with origin server mark, the origin server mark in the step A
Know for distinguishing application request source；
The step B specifically includes following sub-step:
B1, distribution of services module obtain application request；
Application request source is sent to CPU by the network interface card of each calculate node and calculates core by B2, distribution of services module；
B3, CPU calculate core and indicate to compare by application request source and origin server, and comparing result is identical CPU calculating
Core passes through network interface card feedback result to distribution of services module；
B4, distribution of services module are fed back according to above-mentioned steps B3 is sent to corresponding network interface card for the request content of application request.
7. computing cluster control method in a kind of single machine according to claim 6, which is characterized in that the step B is specific
Are as follows: distribution of services module obtains application request, and request content is sent to accordingly after load balancing or HTTP scheduling
Calculate node.
8. computing cluster control method in a kind of single machine according to claim 6, it is characterised in that: every in the step A
The CPU of a calculate node calculates core and is additionally provided with dispatching priority mark, the step B4 specifically:
If having multiple CPU to calculate core in step B3 passes through network interface card feedback result to distribution of services module, distribution of services module
It is identified according to above-mentioned dispatching priority and the request content of application request is sent to corresponding network interface card.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610883006.7A CN106528488B (en) | 2016-10-09 | 2016-10-09 | Computing cluster system and control method in a kind of single machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610883006.7A CN106528488B (en) | 2016-10-09 | 2016-10-09 | Computing cluster system and control method in a kind of single machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106528488A CN106528488A (en) | 2017-03-22 |
| CN106528488B true CN106528488B (en) | 2019-07-16 |
Family
ID=58331243
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610883006.7A Active CN106528488B (en) | 2016-10-09 | 2016-10-09 | Computing cluster system and control method in a kind of single machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106528488B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101631139A (en) * | 2009-05-19 | 2010-01-20 | 华耀环宇科技（北京）有限公司 | Load balancing software architecture based on multi-core platform and method therefor |
| CN102999571A (en) * | 2012-11-09 | 2013-03-27 | 深圳市宜搜科技发展有限公司 | Realizing method for multiple nodes of single computer in cluster |
| CN103634344A (en) * | 2012-08-23 | 2014-03-12 | 阿里巴巴集团控股有限公司 | Method and apparatus for unit operation multiple MySQL database examples |
-
2016
- 2016-10-09 CN CN201610883006.7A patent/CN106528488B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101631139A (en) * | 2009-05-19 | 2010-01-20 | 华耀环宇科技（北京）有限公司 | Load balancing software architecture based on multi-core platform and method therefor |
| CN103634344A (en) * | 2012-08-23 | 2014-03-12 | 阿里巴巴集团控股有限公司 | Method and apparatus for unit operation multiple MySQL database examples |
| CN102999571A (en) * | 2012-11-09 | 2013-03-27 | 深圳市宜搜科技发展有限公司 | Realizing method for multiple nodes of single computer in cluster |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106528488A (en) | 2017-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210289030A1 (en) | Methods, systems and devices for parallel network interface data structures with differential data storage and processing service capabilities | |
| US10412021B2 (en) | Optimizing placement of virtual machines | |
| US11573831B2 (en) | Optimizing resource usage in distributed computing environments by dynamically adjusting resource unit size | |
| US9633054B2 (en) | Providing a database as a service in a multi-tenant environment | |
| US9882985B1 (en) | Data storage path optimization for internet of things computing system | |
| Tang et al. | The impact of data replication on job scheduling performance in the data grid | |
| JP2018088293A (en) | Database system providing single tenant environment and a plurality of tenant environments | |
| US9998531B2 (en) | Computer-based, balanced provisioning and optimization of data transfer resources for products and services | |
| WO2015096681A1 (en) | Method and device for processing service access | |
| CN103338252B (en) | Realizing method of distributed database concurrence storage virtual request mechanism | |
| US9161064B2 (en) | Auto-scaling management of web content | |
| US10776141B2 (en) | Directed placement for request instances | |
| US8898422B2 (en) | Workload-aware distributed data processing apparatus and method for processing large data based on hardware acceleration | |
| JP2017514243A (en) | Coordinated admission control for network accessible block storage | |
| CN110830604B (en) | DNS scheduling method and device | |
| Hu et al. | Towards" full containerization" in containerized network function virtualization | |
| Ma et al. | vLocality: Revisiting data locality for MapReduce in virtualized clouds | |
| WO2024016596A1 (en) | Container cluster scheduling method and apparatus, device, and storage medium | |
| WO2021258512A1 (en) | Data aggregation processing apparatus and method, and storage medium | |
| Wu et al. | iShare: Balancing I/O performance isolation and disk I/O efficiency in virtualized environments | |
| CN106528488B (en) | Computing cluster system and control method in a kind of single machine | |
| CN107239341B (en) | Resource conversion method and system and virtual machine resource scheduling system | |
| US10630767B1 (en) | Hardware grouping based computing resource allocation | |
| Yu et al. | Performance studies of a websphere application, trade, in scale-out and scale-up environments | |
| Han et al. | A locality live migration strategy based on docker containers |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |