CN114979131B - Cloud computing-oriented communication method and device for labeled von neumann architecture - Google Patents
Cloud computing-oriented communication method and device for labeled von neumann architecture Download PDFInfo
- Publication number
- CN114979131B CN114979131B CN202210360889.9A CN202210360889A CN114979131B CN 114979131 B CN114979131 B CN 114979131B CN 202210360889 A CN202210360889 A CN 202210360889A CN 114979131 B CN114979131 B CN 114979131B
- Authority
- CN
- China
- Prior art keywords
- request
- hardware
- resources
- resource
- tag
- 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
- 238000000034 method Methods 0.000 title claims abstract description 117
- 238000004891 communication Methods 0.000 title claims abstract description 50
- 230000008569 process Effects 0.000 claims abstract description 91
- 238000002955 isolation Methods 0.000 claims abstract description 19
- 238000012545 processing Methods 0.000 claims abstract description 17
- 230000007246 mechanism Effects 0.000 claims description 42
- 238000002372 labelling Methods 0.000 claims description 31
- 230000004069 differentiation Effects 0.000 claims description 5
- 230000001902 propagating effect Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 3
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000012913 prioritisation Methods 0.000 description 2
- 238000013468 resource allocation Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Storage Device Security (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a cloud computing-oriented communication method and a cloud computing-oriented communication device for a labeled von neumann system structure. The control plane is added between the request and the hardware resource, the label in the request can be analyzed, the label can be set for the bottom hardware according to the requirement, and meanwhile, the control planes on different computers can communicate, so that the hardware resource can be intensively and jointly called, and the call under the cloud service scene can be realized. Through the structure provided by the invention, the hardware with the tag can execute the processing strategy defined by the application layer and aiming at the tag, so that the on-demand allocation and performance isolation of shared hardware resources are realized, the difficult problem faced by the current cloud computing is solved, the resource utilization rate is improved on the premise of guaranteeing the application QoS, and the interference among the process shared resources is reduced.
Description
Technical Field
The invention relates to a cloud computing-oriented communication method and device of a labeled von neumann architecture.
Background
Cloud computing has the advantages of high performance, service, elasticity and flexibility, environmental friendliness and the like, and has become a widely-used novel IT infrastructure. However, the resource utilization of cloud data centers is very low. The CPU utilization of most data centers worldwide is about 6% -12%. Even the largest public cloud AmazonEC in the world has a CPU utilization of only 7% -17%, which causes a huge loss to IT companies. Recently, both industry and academia have attempted to develop a shared cloud that co-locates different workloads on the same underlying hardware to improve resource utilization while guaranteeing user experience. The shared cloud brings great challenges to cloud providers, including long-tail delay of service, low utilization rate of resources, high interference among hardware resources and the like.
Von neumann architecture is a classical model of a stored program digital computer, proposed by math and physicist john von neumann in 1945. It mainly comprises five parts, a processing unit comprising an arithmetic logic unit and a processor register, a control unit comprising an instruction register and a program counter, a memory for storing data and instructions, external mass storage and input and output mechanisms.
Currently, there is a multi-layered network structure of an extended von neumann structure applied in the internet of things, which divides home or enterprise devices into multi-layered network structures according to device attributes and communication types; the multi-layer network structure comprises an input device network, an output device network and a calculation and control network, wherein the multi-layer network layer is divided, device communication is limited between adjacent layers, communication bandwidth is not shared among devices, the input device network layer is connected with the calculation and control network layer through a cross switch, the output device network layer is connected with the calculation and control network layer through the cross switch, communication among devices is not limited to a common shared wireless network, and high-definition video and other high-speed information are transmitted in parallel among devices.
However, the prior art has a plurality of defects, and mainly comprises the following three parts: long tail delay, low utilization and high interference.
Long tail delay: tail delay was first proposed by google researchers and has immediately attracted a great deal of attention. They found that the rare high delays in a single component may lead to overall service performance on a large scale. The variability of the delay profile of the individual components is scaled up at the service level. To guarantee a user experience in a shared cloud, a first challenge is to suppress delay variability. However, delay variability may be caused by a variety of reasons, including concurrent locks within applications, unordered resource sharing in cache and memory bandwidth, queuing at various resource levels, and so forth.
Low utilization rate: to reduce the delay variation, cloud providers either hire experienced engineers and struggle to carefully optimize their cloud applications, or utilize oversupply to ensure resource utilization by the applications. Both of these methods are costly but have little effect. Meanwhile, over-configuration may result in lower data center resource utilization. The shared cloud encounters a key tradeoff between low utilization and tail latency.
High interference: the basic idea of sharing a cloud is to put different workloads on the same physical server. Different applications sharing the same hardware resources may interfere with each other in performance. With the development of distributed processing frameworks, the execution time of a single task is reduced to hundreds of milliseconds, which makes performance interference more severe and difficult to locate.
Disclosure of Invention
The embodiment of the invention provides a cloud computing-oriented labeled von neumann architecture communication method and device, which are used for at least solving the technical problem of low cloud resource utilization rate in the prior art.
According to an embodiment of the present invention, there is provided a cloud computing-oriented, labeled von neumann architecture communication method, including the steps of:
tagging is performed on a request of a process to access a resource;
adding a programmable interface based on a label mechanism on the hardware and the bus part, and carrying out labeling treatment on the hardware and the bus part;
Adding a control plane between the request and the hardware resource, wherein the control plane can analyze the label in the request and set the label for the bottom hardware according to the requirement; the control planes are distributed on different computers for communication, and hardware resources can be collectively and jointly called.
Further, the computer is configured with a classical von neumann architecture, and the tag mechanism is capable of adding corresponding tags to the request packet, including three tags, distinguishing, quarantining, and prioritizing, setting the priority and exclusivity of the access request.
Further, the distinguishing tag is a D attribute tag mechanism, in the tagged von neumann structure, the tag will be attached to all data access requests, and is used for identifying which application or application category the request originates from, and as the data access request propagates in the whole computer system, the bus and the shared hardware component distinguish the requests of different applications or application categories by checking the tag of the data access request, so as to support distinguishing D attributes;
the isolation tag is an I attribute tag mechanism, and the bus and the shared hardware component isolate the space resources accessed by the request on the basis of source discrimination of the data access request, so that interference caused by sharing conflict of the space resources is relieved or eliminated, and the isolation I attribute is supported;
the priority tag is a P attribute tag, and the bus and the shared hardware component prioritize the performance resources used by the request on the basis of source differentiation of the data access request, so that interference caused by sharing conflict of the performance resources is relieved or eliminated, and the prioritized P attribute is supported.
Further, the D attribute tag is represented by a three-bit binary number, and the tag confirms which application the request originates from, so that the resource required by the request is identified, and accordingly, the hardware resource required to be accessed by the request is selected;
Setting a parameter in the I attribute tag to be 0 or 1, when the system identifies 0, the process occupies resources without isolation, when the system identifies 1, the resources need to be isolated, remarks are requested, and when the resources are to be accessed, the access rights of other requests to the resources are closed until the resources are completely used up;
In the P attribute tag, the priority of the process is confirmed, the priority of the real-time process is higher than that of the normal process, and the normal process uses the nice value to determine the priority.
Further, the nice value is a specified numerical area, and a larger nice value means a lower priority, and the time for which the ordinary request occupies the resource is determined by the nice value.
Further, a request of a process to access a bottom hardware resource first enters an integrated control plane, and the control plane identifies DIP attributes in the request, specifically:
firstly, identifying the attribute D, and confirming the type of the resource which is required to be accessed;
Secondly, the I attribute is identified, when the process is identified to need to isolate the resource, remarks are requested, and when the resource is to be accessed, the access rights of other requests to the resource are closed until the resource is released after being used up;
And finally, identifying the P attribute, judging whether the P attribute is a real-time process, if yes, directly accessing the P attribute and the idle resource meeting the requirement, if not, entering the forefront of a non-real-time process of the waiting queue, if not, entering the waiting queue, and determining the time of occupying the resource according to the priority level confirmed by the nice value.
Further, in a cloud computing scenario, on a cluster of computers employing a tagged von neumann architecture, process tags and hardware tags of different computers can each identify matches, and communication is performed between multiple control planes.
According to another embodiment of the present invention, there is provided a cloud computing-oriented, labeled von neumann architecture communications device, comprising:
The request labeling unit is used for labeling the request of the process for accessing the resource;
the programmable interface adding unit is used for adding a programmable interface based on a label mechanism on the hardware and the bus part and carrying out label processing on the hardware and the bus part;
The control plane adding unit is used for adding a control plane between the request and the hardware resource, wherein the control plane can analyze the label in the request and set the label for the bottom hardware according to the requirement; the control planes are distributed on different computers for communication, and hardware resources can be collectively and jointly called.
A storage medium storing a program file capable of implementing any one of the cloud computing oriented tagged von neumann architecture communication methods described above.
A processor for running a program, wherein the program when running performs any of the cloud computing oriented tagged von neumann architecture communications methods described above.
According to the cloud computing-oriented labeled von neumann system structure communication method and device, a labeling mechanism is added to communication among all parts of the von neumann structure, firstly, requests of processes for accessing resources are labeled, and secondly, programmable interfaces based on the labeling mechanism are added to hardware and bus parts, so that the hardware parts can also carry out labeling processing. The control plane is added between the request and the hardware resource, the label in the request can be analyzed, the label can be set for the bottom hardware according to the requirement, and meanwhile, the control planes on different computers can communicate, so that the hardware resource can be intensively and jointly called, and the call under the cloud service scene can be realized. Through the structure provided by the invention, the hardware with the tag can execute the processing strategy defined by the application layer and aiming at the tag, so that the on-demand allocation and performance isolation of shared hardware resources are realized, the difficult problem faced by the current cloud computing is solved, the resource utilization rate is improved on the premise of guaranteeing the application QoS, and the interference among the process shared resources is reduced.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:
FIG. 1 is a diagram of a tagged von Neumann structure in the present invention;
FIG. 2 is a diagram of a tagged von Neumann diagram under multiple devices in accordance with the present invention;
FIG. 3 is a diagram of the control plane identification process for a requested tag in the present invention;
Fig. 4 is a diagram of a priority identification process according to the present invention.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
According to an embodiment of the present invention, there is provided a cloud computing-oriented, labeled von neumann architecture communication method, including the steps of:
tagging is performed on a request of a process to access a resource;
adding a programmable interface based on a label mechanism on the hardware and the bus part, and carrying out labeling treatment on the hardware and the bus part;
Adding a control plane between the request and the hardware resource, wherein the control plane can analyze the label in the request and set the label for the bottom hardware according to the requirement; the control planes are distributed on different computers for communication, and hardware resources can be collectively and jointly called.
According to the cloud computing-oriented labeled von neumann system structure communication method, a labeling mechanism is added to communication among all parts of the von neumann structure, firstly, requests of processes for accessing resources are labeled, and secondly, programmable interfaces based on the labeling mechanism are added to hardware and bus parts, so that the hardware parts can also carry out labeling processing. The control plane is added between the request and the hardware resource, the label in the request can be analyzed, the label can be set for the bottom hardware according to the requirement, and meanwhile, the control planes on different computers can communicate, so that the hardware resource can be intensively and jointly called, and the call under the cloud service scene can be realized. Through the structure provided by the invention, the hardware with the tag can execute the processing strategy defined by the application layer and aiming at the tag, so that the on-demand allocation and performance isolation of shared hardware resources are realized, the difficult problem faced by the current cloud computing is solved, the resource utilization rate is improved on the premise of guaranteeing the application QoS, and the interference among the process shared resources is reduced.
Wherein the computer is configured with a classical von neumann architecture, the tag mechanism is capable of adding corresponding tags to the request packet, including three tags, distinguishing, quarantining, and prioritizing, and setting the priority and exclusivity of the access request.
The method comprises the steps of selecting a data access request from a data access system, wherein the data access request is transmitted to a bus and shared hardware components, and the bus and shared hardware components are used for identifying the data access request according to the data access request, wherein the identification tag is a D attribute tag mechanism, and in the labeled von neumann structure, the tag is attached to all data access requests and is used for identifying which application or application category the request originates from, and the bus and shared hardware components are used for identifying the requests of different applications or application categories by checking the tags of the data access request together with the data access request in the whole computer system, so that D attribute identification is supported;
the isolation tag is an I attribute tag mechanism, and the bus and the shared hardware component isolate the space resources accessed by the request on the basis of source discrimination of the data access request, so that interference caused by sharing conflict of the space resources is relieved or eliminated, and the isolation I attribute is supported;
the priority tag is a P attribute tag, and the bus and the shared hardware component prioritize the performance resources used by the request on the basis of source differentiation of the data access request, so that interference caused by sharing conflict of the performance resources is relieved or eliminated, and the prioritized P attribute is supported.
The D attribute tag is represented by a three-bit binary number, and the tag confirms which application the request originates from, so that the resource required by the request is identified, and the hardware resource required to be accessed by the request is selected according to the identification;
Setting a parameter in the I attribute tag to be 0 or 1, when the system identifies 0, the process occupies resources without isolation, when the system identifies 1, the resources need to be isolated, remarks are requested, and when the resources are to be accessed, the access rights of other requests to the resources are closed until the resources are completely used up;
In the P attribute tag, the priority of the process is confirmed, the priority of the real-time process is higher than that of the normal process, and the normal process uses the nice value to determine the priority.
Where the nice value is a specified numerical area, a larger nice value means a lower priority, and the time for which the ordinary request occupies the resource is determined by the nice value.
The request of the process for accessing the bottom hardware resource firstly enters an integrated control plane, and the control plane identifies the DIP attribute in the request, specifically:
firstly, identifying the attribute D, and confirming the type of the resource which is required to be accessed;
Secondly, the I attribute is identified, when the process is identified to need to isolate the resource, remarks are requested, and when the resource is to be accessed, the access rights of other requests to the resource are closed until the resource is released after being used up;
And finally, identifying the P attribute, judging whether the P attribute is a real-time process, if yes, directly accessing the P attribute and the idle resource meeting the requirement, if not, entering the forefront of a non-real-time process of the waiting queue, if not, entering the waiting queue, and determining the time of occupying the resource according to the priority level confirmed by the nice value.
In a cloud computing scenario, on a cluster formed by computers adopting a labeled von neumann architecture, process labels and hardware labels of different computers can be identified and matched, and a plurality of control planes can be communicated.
The cloud computing oriented, labeled von neumann architecture communication method of the present invention is described in detail below with specific embodiments:
The invention aims to add labels on each communication link and hardware resources on the basis of a classical von neumann structure, so that the speed of accessing the resources by a process is increased, the requirements and the resources are more pertinently corresponding, the running time of tasks is reduced, the utilization rate of computer resources is improved, the mutual interference of application programs sharing the resources is reduced, and the problem of concurrent locking possibly occurring in part originally is reduced, thereby reducing long-tail delay. The invention aims at adding labels to the process from the input equipment, and utilizing the control center to control, judge and adjust the priority allocation of the bottom hardware resources, thereby realizing the top-down label-based resource allocation, and allocating different resources to the process according to the information provided by the labels to solve the problem, and further improving the utilization rate of the computer resources.
The invention provides a cloud service-oriented labeled von neumann architecture, wherein a labeling mechanism is added to communication among all parts of the von neumann architecture, firstly, a request of a process for accessing resources is labeled, and secondly, a programmable interface based on the labeling mechanism is added to hardware and a bus part, so that the hardware part can also carry out labeling processing. The control plane is added between the request and the hardware resource, the label in the request can be analyzed, the label can be set for the bottom hardware according to the requirement, and meanwhile, the control planes on different computers can communicate, so that the hardware resource can be intensively and jointly called, and the call under the cloud service scene can be realized. Through the structure provided by the invention, the hardware with the tag can execute the processing strategy defined by the application layer and aiming at the tag, so that the on-demand allocation and performance isolation of shared hardware resources are realized, the difficult problem faced by the current cloud computing is solved, the resource utilization rate is improved on the premise of guaranteeing the application QoS, and the interference among the process shared resources is reduced.
1-2, The tag mechanism can add corresponding tags to the request data packet on a computer with a classical von Neumann architecture, including three types of tags of distinguishing, isolating and prioritizing, set the priority and the monopolizing of the access request, and simultaneously add a set of programmable interfaces based on the mechanism, so that when the bus and the shared hardware component support the capability of distinguishing, isolating and prioritizing, the upper layer process can call computer resources more quickly and conveniently when the access request exists, a control plane is added between the request and the hardware resources, the tag in the request can be analyzed, the tag can be set for the hardware of the bottom layer according to the requirement, and meanwhile, the control planes on different computers can communicate, thereby the hardware resources can be intensively and jointly called, and the call under the cloud service scene can be realized. The mechanism can reduce interference caused by resource competition in the computer system, communication between the access request and the hardware is completed through the control plane, and under the cloud service scene, the control planes of different machines can communicate with each other and access the hardware resource in a parallel manner.
The label mechanism of the invention is mainly divided into three types, namely a D attribute, an I attribute and a P attribute.
D attribute tagging mechanism. In a tagged von neumann architecture, tags will be attached to all data access requests to identify from which application (or class of application) the request originated and propagate throughout the computer system with the data access request. In this way, the bus and shared hardware components can support distinguishing attributes (D attributes) by examining the tags of data access requests to distinguish between requests of different applications (or application classes).
The D attribute tag is mainly represented by a three-bit binary number, and the tag confirms from which application the request originates, so as to identify what bandwidth, computing unit, memory and I/O are needed for the resource required by the request, and accordingly, the hardware resource required to be accessed by the request is selected.
I attribute labeling mechanism. The bus and the shared hardware component can isolate the space resources (such as a cache, a memory address space and the like) accessed by the request on the basis of source distinguishing of the data access request, and alleviate or eliminate the interference caused by the sharing conflict of the space resources, so that the isolation attribute (I attribute) is supported.
And setting a parameter in the I attribute tag, wherein the parameter can be set to be 0 or 1, when the system is identified as 0, the process occupies resources without isolation, when the system is identified as1, the resources need to be isolated, remarks are requested, and when the resources are to be accessed, the access rights of other requests to the resources are closed until the resources are completely used.
P attribute labeling mechanism. The bus and shared hardware components may prioritize performance resources (e.g., queues, bandwidth, etc.) used by requests based on source differentiation of data access requests, and mitigate or eliminate interference due to shared conflicts of performance resources, thereby supporting a prioritization attribute (P-attribute).
In the P attribute tag, first, the priority of the process is confirmed. In the invention, the priority of the real-time process is absolutely higher than that of the common process. The normal process uses the nice value to determine priority, where a nice value is a specified numerical region, a larger nice value meaning a lower priority, and the time at which the normal request occupies resources is determined by the nice value.
Referring to fig. 3, the present invention employs an integrated control plane into which a request for a process to access a bottom hardware resource first enters, the control plane identifying DIP attributes in the request. Firstly, identifying the attribute D, and confirming the type of the resource which is required to be accessed; and secondly, identifying the I attribute, requesting remarks when the process is identified to isolate the resource, and closing the access rights of other requests to the resource when the resource is to be accessed until the resource is released after the use. Referring to fig. 4, finally, the P attribute is identified, and whether the process is a real-time process is judged, if yes, the process is directly accessed, and if the idle resource meeting the requirement exists. If there are not enough resources, the front end of the non-real-time process of the waiting queue is entered. If the process is not a real-time process, entering a waiting queue, and determining the time of occupying resources according to the priority level confirmed by the nice value.
The programmable interface based on the mechanism is added on hardware, so that the system has a labeling function.
On a cluster formed by computers adopting a labeled von neumann architecture, the process labels of different computers can be identified and matched with the hardware labels, and meanwhile, the control planes can be communicated, so that the resource utilization rate can be improved in a cloud computing scene.
Example 2
According to another embodiment of the present invention, there is provided a cloud computing-oriented, labeled von neumann architecture communications device, comprising:
The request labeling unit is used for labeling the request of the process for accessing the resource;
the programmable interface adding unit is used for adding a programmable interface based on a label mechanism on the hardware and the bus part and carrying out label processing on the hardware and the bus part;
The control plane adding unit is used for adding a control plane between the request and the hardware resource, wherein the control plane can analyze the label in the request and set the label for the bottom hardware according to the requirement; the control planes are distributed on different computers for communication, and hardware resources can be collectively and jointly called.
According to the cloud computing-oriented labeled von neumann system structure communication device, a labeling mechanism is added to communication among all parts of the von neumann structure, firstly, requests of processes for accessing resources are labeled, and secondly, programmable interfaces based on the labeling mechanism are added to hardware and bus parts, so that the hardware parts can also carry out labeling processing. The control plane is added between the request and the hardware resource, the label in the request can be analyzed, the label can be set for the bottom hardware according to the requirement, and meanwhile, the control planes on different computers can communicate, so that the hardware resource can be intensively and jointly called, and the call under the cloud service scene can be realized. Through the structure provided by the invention, the hardware with the tag can execute the processing strategy defined by the application layer and aiming at the tag, so that the on-demand allocation and performance isolation of shared hardware resources are realized, the difficult problem faced by the current cloud computing is solved, the resource utilization rate is improved on the premise of guaranteeing the application QoS, and the interference among the process shared resources is reduced.
The cloud computing oriented, labeled von neumann architecture communication device of the present invention is described in detail below with specific embodiments:
The invention aims to add labels on each communication link and hardware resources on the basis of a classical von neumann structure, so that the speed of accessing the resources by a process is increased, the requirements and the resources are more pertinently corresponding, the running time of tasks is reduced, the utilization rate of computer resources is improved, the mutual interference of application programs sharing the resources is reduced, and the problem of concurrent locking possibly occurring in part originally is reduced, thereby reducing long-tail delay. The invention aims at adding labels to the process from the input equipment, and utilizing the control center to control, judge and adjust the priority allocation of the bottom hardware resources, thereby realizing the top-down label-based resource allocation, and allocating different resources to the process according to the information provided by the labels to solve the problem, and further improving the utilization rate of the computer resources.
The invention provides a cloud service-oriented labeled von neumann architecture, wherein a labeling mechanism is added to communication among all parts of the von neumann architecture, firstly, a request of a process for accessing resources is labeled, and secondly, a programmable interface based on the labeling mechanism is added to hardware and a bus part, so that the hardware part can also carry out labeling processing. The control plane is added between the request and the hardware resource, the label in the request can be analyzed, the label can be set for the bottom hardware according to the requirement, and meanwhile, the control planes on different computers can communicate, so that the hardware resource can be intensively and jointly called, and the call under the cloud service scene can be realized. Through the structure provided by the invention, the hardware with the tag can execute the processing strategy defined by the application layer and aiming at the tag, so that the on-demand allocation and performance isolation of shared hardware resources are realized, the difficult problem faced by the current cloud computing is solved, the resource utilization rate is improved on the premise of guaranteeing the application QoS, and the interference among the process shared resources is reduced.
1-2, The tag mechanism can add corresponding tags to the request data packet on a computer with a classical von Neumann architecture, including three types of tags of distinguishing, isolating and prioritizing, set the priority and the monopolizing of the access request, and simultaneously add a set of programmable interfaces based on the mechanism, so that when the bus and the shared hardware component support the capability of distinguishing, isolating and prioritizing, the upper layer process can call computer resources more quickly and conveniently when the access request exists, a control plane is added between the request and the hardware resources, the tag in the request can be analyzed, the tag can be set for the hardware of the bottom layer according to the requirement, and meanwhile, the control planes on different computers can communicate, thereby the hardware resources can be intensively and jointly called, and the call under the cloud service scene can be realized. The mechanism can reduce interference caused by resource competition in the computer system, communication between the access request and the hardware is completed through the control plane, and under the cloud service scene, the control planes of different machines can communicate with each other and access the hardware resource in a parallel manner.
The label mechanism of the invention is mainly divided into three types, namely a D attribute, an I attribute and a P attribute.
D attribute tagging mechanism. In a tagged von neumann architecture, tags will be attached to all data access requests to identify from which application (or class of application) the request originated and propagate throughout the computer system with the data access request. In this way, the bus and shared hardware components can support distinguishing attributes (D attributes) by examining the tags of data access requests to distinguish between requests of different applications (or application classes).
The D attribute tag is mainly represented by a three-bit binary number, and the tag confirms from which application the request originates, so as to identify what bandwidth, computing unit, memory and I/O are needed for the resource required by the request, and accordingly, the hardware resource required to be accessed by the request is selected.
I attribute labeling mechanism. The bus and the shared hardware component can isolate the space resources (such as a cache, a memory address space and the like) accessed by the request on the basis of source distinguishing of the data access request, and alleviate or eliminate the interference caused by the sharing conflict of the space resources, so that the isolation attribute (I attribute) is supported.
And setting a parameter in the I attribute tag, wherein the parameter can be set to be 0 or 1, when the system is identified as 0, the process occupies resources without isolation, when the system is identified as1, the resources need to be isolated, remarks are requested, and when the resources are to be accessed, the access rights of other requests to the resources are closed until the resources are completely used.
P attribute labeling mechanism. The bus and shared hardware components may prioritize performance resources (e.g., queues, bandwidth, etc.) used by requests based on source differentiation of data access requests, and mitigate or eliminate interference due to shared conflicts of performance resources, thereby supporting a prioritization attribute (P-attribute).
In the P attribute tag, first, the priority of the process is confirmed. In the invention, the priority of the real-time process is absolutely higher than that of the common process. The normal process uses the nice value to determine priority, where a nice value is a specified numerical region, a larger nice value meaning a lower priority, and the time at which the normal request occupies resources is determined by the nice value.
Referring to fig. 3, the present invention employs an integrated control plane into which a request for a process to access a bottom hardware resource first enters, the control plane identifying DIP attributes in the request. Firstly, identifying the attribute D, and confirming the type of the resource which is required to be accessed; and secondly, identifying the I attribute, requesting remarks when the process is identified to isolate the resource, and closing the access rights of other requests to the resource when the resource is to be accessed until the resource is released after the use. Referring to fig. 4, finally, the P attribute is identified, and whether the process is a real-time process is judged, if yes, the process is directly accessed, and if the idle resource meeting the requirement exists. If there are not enough resources, the front end of the non-real-time process of the waiting queue is entered. If the process is not a real-time process, entering a waiting queue, and determining the time of occupying resources according to the priority level confirmed by the nice value.
The programmable interface based on the mechanism is added on hardware, so that the system has a labeling function.
On a cluster formed by computers adopting a labeled von neumann architecture, the process labels of different computers can be identified and matched with the hardware labels, and meanwhile, the control planes can be communicated, so that the resource utilization rate can be improved in a cloud computing scene.
Example 3
A storage medium storing a program file capable of implementing any one of the cloud computing oriented tagged von neumann architecture communication methods described above.
Example 4
A processor for running a program, wherein the program when running performs any of the cloud computing oriented tagged von neumann architecture communications methods described above.
The invention provides a labeling method, wherein labels are added in a software process data packet to distinguish different requirements of a process, and a control plane is adopted to coordinate computer resources, so that different allocation strategies are provided for different requirements of the process; the control plane capable of communicating among different computers is provided, and hardware resources on different computers can be jointly accessed and managed through the control plane communication, so that the resource utilization rate in a cloud service scene is improved.
The method has the technical advantages that: the resource utilization rate of the computer in the cloud service scene can be improved; the mutual interference between processes sharing hardware resources can be reduced. Experiments prove that the invention is efficient and feasible.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
In the foregoing embodiments of the present invention, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.
In the several embodiments provided in the present application, it should be understood that the disclosed technology may be implemented in other manners. The system embodiments described above are merely exemplary, and for example, the division of units may be a logic function division, and there may be another division manner in actual implementation, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method of the various embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a read-only memory (ROM), a random access memory (RAM, random Access Memory), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The foregoing is merely a preferred embodiment of the present invention and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present invention, which are intended to be comprehended within the scope of the present invention.
Claims (7)
1. A cloud computing-oriented method of communication of a tagged von neumann architecture, comprising the steps of:
tagging is performed on a request of a process to access a resource;
adding a programmable interface based on a label mechanism on the hardware and the bus part, and carrying out labeling treatment on the hardware and the bus part;
Adding a control plane between the request and the hardware resource, wherein the control plane can analyze the label in the request and set the label for the bottom hardware according to the requirement; the control planes are distributed on different computers for communication, so that hardware resources can be intensively and jointly called;
The computer is configured to have a classical von neumann architecture, and the tag mechanism adds corresponding tags to the request packet, including distinguishing tags, quarantine tags, and priority tags, setting priority and exclusivity of the access request;
The distinguishing tag is a D attribute tag mechanism, in the tagged von Neumann structure, the tag is attached to all data access requests and is used for identifying which application or application category the request originates from, and the bus and the shared hardware component distinguish the requests of different applications or application categories by checking the tag of the data access request together with the data access request propagating in the whole computer system, so as to support distinguishing D attributes;
the isolation tag is an I attribute tag mechanism, and the bus and the shared hardware component isolate the space resources accessed by the request on the basis of source discrimination of the data access request, so that interference caused by sharing conflict of the space resources is relieved or eliminated, and the isolation I attribute is supported;
the priority tag is a P attribute tag, and the bus and the shared hardware component prioritize the performance resources used by the request on the basis of source differentiation of the data access request, so that interference caused by sharing conflict of the performance resources is relieved or eliminated, and the prioritized P attribute is supported.
2. The cloud computing-oriented tagged von neumann architecture communication method of claim 1, wherein the D-attribute tag is represented by a three-bit binary number, identifying from which application the request originated by the tag, thereby identifying the resources required by the request, whereby the hardware resources that the request needs to access are selected;
Setting a parameter in the I attribute tag to be 0 or 1, when the system identifies 0, the process occupies resources without isolation, when the system identifies 1, the resources need to be isolated, remarks are requested, and when the resources are to be accessed, the access rights of other requests to the resources are closed until the resources are completely used up;
In the P attribute tag, the priority of the process is confirmed, the priority of the real-time process is higher than that of the normal process, and the normal process uses the nice value to determine the priority.
3. The cloud computing-oriented tagged von neumann architecture communication method of claim 2, wherein the nice value is a specified numerical region, a larger nice value implying a lower priority, and the time for a common request to occupy a resource is determined by the nice value.
4. The cloud computing-oriented tagged von neumann architecture communication method of claim 2, characterized in that a request for a process to access a bottom hardware resource first enters an integrated control plane that identifies DIP attributes in the request, in particular:
firstly, identifying the attribute D, and confirming the type of the resource which is required to be accessed;
Secondly, the I attribute is identified, when the process is identified to need to isolate the resource, remarks are requested, and when the resource is to be accessed, the access rights of other requests to the resource are closed until the resource is released after being used up;
And finally, identifying the P attribute, judging whether the P attribute is a real-time process, if yes, directly accessing the P attribute and the idle resource meeting the requirement, if not, entering the forefront of a non-real-time process of the waiting queue, if not, entering the waiting queue, and determining the time of occupying the resource according to the priority level confirmed by the nice value.
5. The cloud computing-oriented tagged von neumann architecture communication method of claim 1, wherein, in a cloud computing scenario, process tags and hardware tags of different computers can be identified and matched on a cluster formed by computers adopting a tagged von neumann architecture, and communication is performed among a plurality of control planes.
6. A cloud computing-oriented tagged von neumann architecture communication device utilizing the cloud computing-oriented tagged von neumann architecture communication method of claim 1, comprising:
The request labeling unit is used for labeling the request of the process for accessing the resource;
the programmable interface adding unit is used for adding a programmable interface based on a label mechanism on the hardware and the bus part and carrying out label processing on the hardware and the bus part;
The control plane adding unit is used for adding a control plane between the request and the hardware resource, wherein the control plane can analyze the label in the request and set the label for the bottom hardware according to the requirement; the control planes are distributed on different computers for communication, and hardware resources can be collectively and jointly called.
7. A storage medium storing program files enabling the cloud computing oriented tagged von neumann architecture communication method of any one of claims 1 to 5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210360889.9A CN114979131B (en) | 2022-04-07 | 2022-04-07 | Cloud computing-oriented communication method and device for labeled von neumann architecture |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210360889.9A CN114979131B (en) | 2022-04-07 | 2022-04-07 | Cloud computing-oriented communication method and device for labeled von neumann architecture |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114979131A CN114979131A (en) | 2022-08-30 |
CN114979131B true CN114979131B (en) | 2024-04-19 |
Family
ID=82978472
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210360889.9A Active CN114979131B (en) | 2022-04-07 | 2022-04-07 | Cloud computing-oriented communication method and device for labeled von neumann architecture |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114979131B (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5706466A (en) * | 1995-01-13 | 1998-01-06 | Vlsi Technology, Inc. | Von Neumann system with harvard processor and instruction buffer |
US5778434A (en) * | 1995-06-07 | 1998-07-07 | Seiko Epson Corporation | System and method for processing multiple requests and out of order returns |
US7050807B1 (en) * | 2000-06-12 | 2006-05-23 | General Dynamics Decision Systems, Inc. | Hardware resource identifier for software-defined communications system |
CN104318171A (en) * | 2014-10-09 | 2015-01-28 | 中国科学院信息工程研究所 | Android privacy data protection method and system based on authority tags |
CN105094983A (en) * | 2014-04-30 | 2015-11-25 | 华为技术有限公司 | Computer, control equipment and data processing method |
CN109101322A (en) * | 2018-08-29 | 2018-12-28 | 山东师范大学 | The safe calculation method of virtualization and system monitored based on paired tag and migration |
CN111897651A (en) * | 2020-07-28 | 2020-11-06 | 华中科技大学 | Memory system resource management method based on tags |
CN114201300A (en) * | 2021-12-14 | 2022-03-18 | 建信金融科技有限责任公司 | Resource scheduling method, resource scheduling apparatus, electronic device, and storage medium |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7797699B2 (en) * | 2004-09-23 | 2010-09-14 | Intel Corporation | Method and apparatus for scheduling virtual machine access to shared resources |
US10586076B2 (en) * | 2015-08-24 | 2020-03-10 | Acronis International Gmbh | System and method for controlling access to OS resources |
US11580239B2 (en) * | 2019-10-22 | 2023-02-14 | Microsoft Technology Licensing, Llc | Controlling access to cloud resources in data using cloud-enabled data tagging and a dynamic access control policy engine |
-
2022
- 2022-04-07 CN CN202210360889.9A patent/CN114979131B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5706466A (en) * | 1995-01-13 | 1998-01-06 | Vlsi Technology, Inc. | Von Neumann system with harvard processor and instruction buffer |
US5778434A (en) * | 1995-06-07 | 1998-07-07 | Seiko Epson Corporation | System and method for processing multiple requests and out of order returns |
US7050807B1 (en) * | 2000-06-12 | 2006-05-23 | General Dynamics Decision Systems, Inc. | Hardware resource identifier for software-defined communications system |
CN105094983A (en) * | 2014-04-30 | 2015-11-25 | 华为技术有限公司 | Computer, control equipment and data processing method |
CN104318171A (en) * | 2014-10-09 | 2015-01-28 | 中国科学院信息工程研究所 | Android privacy data protection method and system based on authority tags |
CN109101322A (en) * | 2018-08-29 | 2018-12-28 | 山东师范大学 | The safe calculation method of virtualization and system monitored based on paired tag and migration |
CN111897651A (en) * | 2020-07-28 | 2020-11-06 | 华中科技大学 | Memory system resource management method based on tags |
CN114201300A (en) * | 2021-12-14 | 2022-03-18 | 建信金融科技有限责任公司 | Resource scheduling method, resource scheduling apparatus, electronic device, and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN114979131A (en) | 2022-08-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105893126B (en) | A kind of method for scheduling task and device | |
US9436510B2 (en) | System and method for managing the interleaved execution of threads | |
US8676976B2 (en) | Microprocessor with software control over allocation of shared resources among multiple virtual servers | |
US8424007B1 (en) | Prioritizing tasks from virtual machines | |
US9448864B2 (en) | Method and apparatus for processing message between processors | |
US20170185452A1 (en) | Apparatus and method for data processing | |
CN111782355B (en) | Cloud computing task scheduling method and system based on mixed load | |
KR20100066458A (en) | Allocating network adapter resources among logical partitions | |
US9507633B2 (en) | Scheduling method and system | |
US7243354B1 (en) | System and method for efficiently processing information in a multithread environment | |
Zhang et al. | Tapping into nfv environment for opportunistic serverless edge function deployment | |
Maruf et al. | Extending resources for avoiding overloads of mixed‐criticality tasks in cyber‐physical systems | |
Mutlu et al. | Parallelism-aware batch scheduling: Enabling high-performance and fair shared memory controllers | |
CN112925616A (en) | Task allocation method and device, storage medium and electronic equipment | |
CN110413210B (en) | Method, apparatus and computer program product for processing data | |
CN114461365A (en) | Process scheduling processing method, device, equipment and storage medium | |
CN111930485B (en) | Job scheduling method based on performance expression | |
CN114979131B (en) | Cloud computing-oriented communication method and device for labeled von neumann architecture | |
CN115878333A (en) | Method, device and equipment for judging consistency between process groups | |
Xu et al. | Optimal construction of virtual networks for cloud-based MapReduce workflows | |
Deshmukh et al. | Avoiding slow running nodes in distributed systems | |
CN113760524A (en) | Task execution method and device | |
CN113296957A (en) | Method and device for dynamically allocating network-on-chip bandwidth | |
US20140298352A1 (en) | Computer with plurality of processors sharing process queue, and process dispatch processing method | |
Huang et al. | EDF‐Adaptive: A New Semipartitioned Scheduling Algorithm for Multiprocessor Real‐Time |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | 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 |