CN109040214B - Service deployment method for enhancing reliability in cloud environment - Google Patents

Service deployment method for enhancing reliability in cloud environment Download PDF

Info

Publication number
CN109040214B
CN109040214B CN201810824345.7A CN201810824345A CN109040214B CN 109040214 B CN109040214 B CN 109040214B CN 201810824345 A CN201810824345 A CN 201810824345A CN 109040214 B CN109040214 B CN 109040214B
Authority
CN
China
Prior art keywords
service
sub
tail
node
head
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
Application number
CN201810824345.7A
Other languages
Chinese (zh)
Other versions
CN109040214A (en
Inventor
周傲
王尚广
孙其博
李静林
原媛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing University of Posts and Telecommunications
Original Assignee
Beijing University of Posts and Telecommunications
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing University of Posts and Telecommunications filed Critical Beijing University of Posts and Telecommunications
Priority to CN201810824345.7A priority Critical patent/CN109040214B/en
Publication of CN109040214A publication Critical patent/CN109040214A/en
Application granted granted Critical
Publication of CN109040214B publication Critical patent/CN109040214B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/142—Network analysis or design using statistical or mathematical methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5041—Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the time relationship between creation and deployment of a service
    • 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
    • H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
    • 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
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
    • 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
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/101—Server selection for load balancing based on network conditions

Abstract

The method comprises the steps of firstly modeling the complex service into a graph model, enabling main nodes of the graph to represent sub-services of the complex service, enabling edges of the graph to represent data transmission rates among the sub-services, then selecting corresponding reliability strategies for the sub-services according to the characteristics of the sub-services, adding key point storage nodes or redundant nodes for the main nodes in the graph model based on the selected reliability strategies, and carrying out corresponding reliability enhancement service deployment.

Description

Service deployment method for enhancing reliability in cloud environment
Technical Field
The application relates to the technical field of computer application, in particular to a service deployment method for enhancing reliability in a cloud environment.
Background
When complex services are deployed in a cloud computing platform, each sub-service of the complex services needs to be deployed into a virtual machine placed above a physical machine, and the complex tasks are executed through cooperation of the virtual machines. Due to the fact that the reliability of the service is reduced due to the fact that the virtual resources are not continuously occupied due to the failure event, measures such as key point storage or virtual machine redundancy placement need to be taken to improve the reliability of the service. The current reliability strategy adopts the same reliability strategy for each sub-service by neglecting the difference of the sub-service calculation characteristics and the difference of different sub-services in the aspect of reliability requirements, so that the optimal reliability guarantee effect is difficult to obtain.
Disclosure of Invention
In view of this, the main objective of the present invention is to provide a service deployment method with enhanced reliability in a cloud environment, so as to minimize additional network resource overhead caused by a reliability policy while ensuring the reliability of a complex service.
The application discloses a service deployment method for enhancing reliability in a cloud environment, which comprises the following steps:
a: modeling the complex service into a graph model, wherein a main node of the graph represents sub-services of the complex service, and edges of the graph represent data transmission rates among the sub-services;
b: traversing each sub-service of the complex service, and selecting key point storage or resource redundancy for each sub-service of the complex service as a reliability strategy of the sub-service according to the characteristics of the sub-service;
c: expanding the graph model based on the selected reliability strategy, and respectively adding a key point storage node or a redundant node for each main node;
d: sequencing all edges connected with the main nodes in the graph model according to the descending order of the data transmission rate, sequentially mapping head and tail nodes of the edges to a physical machine of the cloud data center, distributing virtual machines on the physical machine according to requirements, and deploying sub-services corresponding to the nodes to the virtual machines;
e: sequencing all edges connecting the main nodes and the key point storage nodes in the graph model according to the descending order of the data transmission rate, sequentially mapping tail nodes of the edges to a physical machine of the cloud data center, and distributing storage space for the key points on the physical machine;
f: and sequencing the rest edges in the graph model according to the descending order of the data transmission rate, sequentially mapping the tail nodes to the physical machine of the cloud data center, and distributing the virtual machines for the redundant service nodes on the physical machine.
Preferably, the b comprises:
traversing all sub-services of the complex service, for a currently traversed sub-service siAccording to the time T _ exe(s) required for uninterrupted executioni) Time T _ cp(s) required for one key point storage for the sub-servicei) And an acceptable period T _ interval(s) for storing the key pointsi) The percentage of dead time perc is calculated based on the following formula:
Figure BDA0001742109880000021
then, comparing perc with an acceptability threshold delta, and if perc is greater than delta, selecting resource redundancy for the sub-service as a reliability strategy thereof; if perc is less than or equal to perc, selecting the key point storage for the sub-service as the reliability policy of the sub-service.
Preferably, the a comprises:
a 1: constructing an empty graph model G (V, E) for the complex service, wherein the node set V and the edge set E are empty;
a 2: for complex services S ═ S1,s2...si...smax(s)Each sub-service s ofiAdding a graph node
Figure BDA0001742109880000022
The type of the node is a main node, the attribute of the node comprises the requirements of the sub-service on CPU resources, memory resources and disk resources, and the node is added into a node set V of G;
a 3: for sub-services s with data interactioniAnd sjIs its corresponding node
Figure BDA0001742109880000031
And
Figure BDA0001742109880000032
adding a side
Figure BDA0001742109880000033
And adding the edge to an edge set E of G, wherein the type of the edge is a main edge, and the attribute of the edge is a sub-service siAnd sjThe data transfer rate between.
Preferably, the c comprises:
c 1: traversing and selecting key point storage as a sub-service of the reliability strategy, and thenPre-traversed sub-services siCorresponding main node
Figure BDA0001742109880000034
Adding corresponding key point storage nodes
Figure BDA0001742109880000035
The type of the node is a key point node, and the attribute of the node is a storage space required for storing a key point file
Figure BDA0001742109880000036
Then, it is calculated based on the following formula
Figure BDA0001742109880000037
And
Figure BDA0001742109880000038
data transmission rate of (2):
Figure BDA0001742109880000039
and adding a connection
Figure BDA00017421098800000310
And
Figure BDA00017421098800000311
is not limited by
Figure BDA00017421098800000312
The type of the edge is a storage edge, and the attribute of the edge is
Figure BDA00017421098800000313
And
Figure BDA00017421098800000314
and will be at the data transmission rate of
Figure BDA00017421098800000315
Adding to the graphIn the node set V of the model, will
Figure BDA00017421098800000316
Adding the edge set E into the graph model;
c 2: traversing and selecting resource redundancy as a sub-service of the reliability strategy, and performing the sub-service s of the current traversaliCorresponding main node
Figure BDA00017421098800000330
Adding corresponding redundant nodes
Figure BDA00017421098800000317
And, for any master node in set V
Figure BDA00017421098800000318
If it is with
Figure BDA00017421098800000319
Directly connected with each other, and added with an edge
Figure BDA00017421098800000320
The edge attribute is
Figure BDA00017421098800000321
And
Figure BDA00017421098800000322
the type of edge is a redundant edge, will
Figure BDA00017421098800000323
Added to the set E.
Preferably, d includes:
d 1: adding all main edges in the edge set E of the graph model into a list Ranked _ EM [ ], and sorting the list Ranked _ EM [ ] from high to low according to the data transmission rate;
d2 traversing the list Ranked _ EM [ 2 ]]For the currently traversed edge eiObtaining eiHead (e) ofi) And tail node tail (e)i)；
d3 if head (e)i) And tail (e)i) If the corresponding sub-service is not deployed, searching the sub-network of the data center, and if two physical machines exist
Figure BDA00017421098800000324
And
Figure BDA00017421098800000325
of a sub-network of
Figure BDA00017421098800000329
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure BDA00017421098800000326
is greater than tail (e)i) The resource requirement of (1) is
Figure BDA00017421098800000327
Push head (e)i) Resource demand of (c) allocates virtual machines, head (e)i) The corresponding sub-service is deployed on the virtual machine
Figure BDA0001742109880000041
Push on tail (e)i) Allocate tail (e) to the virtual machinei) Deploying the corresponding sub-service to the virtual machine, and returning to the step d 2;
d4 if no subnet satisfying the condition is found in step d3, searching the cluster of the data center if there are two physical machines
Figure BDA0001742109880000042
And
Figure BDA0001742109880000043
is satisfied by
Figure BDA0001742109880000044
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure BDA0001742109880000045
is greater than tail (e)i) The resource requirement of (1) is
Figure BDA0001742109880000046
Push head (e)i) Resource demand of (c) allocates virtual machines, head (e)i) The corresponding sub-service is deployed on the virtual machine
Figure BDA0001742109880000047
Push on tail (e)i) Allocate tail (e) to the virtual machinei) Deploying the corresponding sub-service to the virtual machine, and returning to the step d 2;
d5 if there are no clusters satisfying the condition at step d4, then two physical machines are sought from the entire data center
Figure BDA0001742109880000048
And
Figure BDA0001742109880000049
satisfy the requirement of
Figure BDA00017421098800000410
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure BDA00017421098800000411
is greater than tail (e)i) Resource requirement of (c), head (e)i) Corresponding sub-service deployment to
Figure BDA00017421098800000412
tail(ei) Corresponding sub-service deployment to
Figure BDA00017421098800000413
Returning to step d 2;
d6 if there is only head (e)i) Not yet deployed to deploy tail (e)i) Is the center of the physical machine, and searches for the residual resource from near to far to be larger than head (e)i) Of (2)Physical machine of source demand on which head (e) is pressedi) Allocating virtual machines to the demand of resources, head (e)i) Deploying the corresponding sub-service on the virtual machine, and returning to the step d 2;
d7 if there is only tail (e)i) Not yet deployed to deploy head (e)i) Is the center of the physical machine, and searches for the residual resource larger than tail (e) from near to fari) The physical machine on which tail (e) is pressedi) Allocating virtual machines to the demand for resources, tail (e)i) The corresponding sub-service is deployed on the virtual machine, and the step d2 is returned.
Preferably, said e includes:
e 1: adding all storage edges in an edge set E of the graph model into a list Ranked _ EC [ ], and sorting the Ranked _ EC [ ] from high to low according to the data transmission rate;
e2 traversing Ranked _ EC 2]For the currently traversed edge eiObtaining head node head (e)i) And tail node tail (e)i)；
e3 with head (e)i) The deployed physical machine is used as a center, and the residual resources are searched from near to far and are larger than tail (e)i) The physical machine of resource demand of (1), the physical machine being allocated as head (e)i) The key point of (2) stores the physical machine.
Preferably, the f includes:
f 1: adding all redundant edges in an edge set E of the graph model into a list Ranked _ ER [ ], and sorting the Ranked _ ER [ ] from high to low according to the data transmission rate;
f2 traversal of Ranked _ ER [ alpha ]]For the currently traversed edge eiObtaining head node head (e)i) And tail node tail (e)i)；
f3 with head (e)i) The deployed physical machine is used as a center, and the residual resources are searched from near to far and are larger than tail (e)i) Physical machine of resource demand on which tail (e) is pressedi) Allocating virtual machines to the demand for resources, tail (e)i) A corresponding sub-service is deployed on the virtual machine.
According to the technical scheme, the service deployment method for enhancing the reliability in the cloud environment, which is provided by the application, models the complex service into a graph model, enables the main nodes of the graph to represent sub-services of the complex service, the edges of the graph to represent the data transmission rate between the sub-services, then selects corresponding reliability strategies for the sub-services according to the characteristics of the sub-services, and adds key point storage nodes or redundant nodes for the main nodes in the graph model based on the selected reliability strategies to perform corresponding service deployment for enhancing the reliability.
Drawings
Fig. 1 is a flowchart illustrating a service deployment method with enhanced reliability in a cloud environment according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is further described in detail below by referring to the accompanying drawings and examples.
In order to solve the problems in the prior art, the application provides a service deployment method for enhancing reliability in a cloud environment, the method selects a corresponding reliability strategy for each sub-service according to the characteristics of each sub-service, and performs service deployment for enhancing reliability while considering the difference of network resource occupation brought by the reliability guarantee of each sub-service so as to minimize the extra network resource occupation brought by the reliability guarantee.
The application discloses a service deployment method for enhancing reliability in a cloud environment, which comprises the following steps:
step 1: the complex service is modeled as a graph model, the main nodes of the graph representing sub-services of the complex service, and the edges of the graph representing data transfer rates between the sub-services.
Step 2: and traversing each sub-service of the complex service, and selecting key point storage or resource redundancy for each sub-service of the complex service as a reliability strategy of the sub-service according to the characteristics of the sub-service of the complex service.
And step 3: and (3) expanding the graph model in the step (1) based on the reliability strategy selected in the step (2), and respectively adding a key point storage node or a redundant node for each main node.
And 4, step 4: sequencing all edges connected with the main nodes in the graph model according to the descending order of the data transmission rate, sequentially mapping head and tail nodes of the edges to a physical machine of the cloud data center, distributing virtual machines on the physical machine according to requirements, and deploying sub-services corresponding to the nodes to the virtual machines.
And 5: and sequencing all edges connecting the main nodes and the key point storage nodes in the graph model according to the descending order of the data transmission rate, mapping tail nodes of the edges to a physical machine of the cloud data center in sequence, and distributing storage space for the key points on the physical machine.
Step 6: and sequencing the rest edges in the graph model according to the descending order of the data transmission rate, sequentially mapping the tail nodes to the physical machine of the cloud data center, and distributing the virtual machines for the redundant service nodes on the physical machine.
And after virtual resources are distributed to all the nodes according to the steps 1 to 6, service deployment with enhanced reliability in the cloud environment is obtained. The method and the device can improve the reliability of the complex service in the cloud environment.
Fig. 1 is a schematic flowchart of a service deployment method for enhancing reliability in a cloud environment according to an embodiment of the present invention, and as shown in fig. 1, the method mainly includes the following steps:
step a: the complex service is modeled as a graph model, the main nodes of the graph representing sub-services of the complex service, and the edges of the graph representing data transfer rates between the sub-services.
One specific method of modeling is listed below:
a 1: and constructing an empty graph model G (V, E) for the complex service, wherein the node set V and the edge set E are empty.
a 2: for complex services S ═ S1,s2...si...smax(s)Each sub-service s ofiAdding a graph node
Figure BDA0001742109880000061
Type of nodeAnd the node is a master node, the attribute of the node comprises the requirements of the sub-service on resources such as CPU (Central processing Unit) resources, memory resources, disk resources and the like, and the node is added into the node set V of G.
a 3: for sub-services s with data interactioniAnd sjIs its corresponding node
Figure BDA0001742109880000071
And
Figure BDA0001742109880000072
adding a side
Figure BDA0001742109880000073
And adding the edge to an edge set E of G, wherein the type of the edge is a main edge, and the attribute of the edge is a sub-service siAnd sjThe data transfer rate between.
Step b: and traversing each sub-service of the complex service, and selecting key point storage or resource redundancy for each sub-service of the complex service as a reliability strategy of the sub-service according to the characteristics of the sub-service of the complex service.
A preferred method for selecting a reliability policy is:
traversing all sub-services of a complex service, for a currently traversed sub-service siAccording to the time T _ exe(s) required for uninterrupted executioni) Time T _ cp(s) required for one key point storage for the sub-servicei) And an acceptable period T _ interval(s) for storing the key pointsi) The percentage of dead time perc is calculated based on the following formula:
Figure BDA0001742109880000074
then, perc is compared with the threshold of acceptability delta, and if perc is greater than delta, resource redundancy is selected for this sub-service as its reliability policy, i.e.: will serve the sub-service siAdd redundancy policy List RD [ 2 ]](ii) a If perc is less than or equal to perc, selecting the key point storage for the sub-service as the reliability policy of the sub-service, namely: will serve the sub-service siAdding a Key PointStorage policy List CP [ 2 ]]。
Here, the redundancy policy list RD [ ] and the key point storage policy list CP [ ] are merely examples, and in practical applications, other ways may also be adopted to record and select resource redundancy or key point storage as sub-services of the reliability policy, respectively.
Step c: and b, according to the reliability strategy selected in the step b, expanding the graph model obtained in the step a, and respectively adding a key point storage node or a redundant node for each main node (namely, the node corresponding to each sub-service) to obtain a new graph model. A preferred implementation is:
c 1: traversal key point storage strategy list CP]For currently traversed sub-service siCorresponding main node
Figure BDA0001742109880000075
Adding corresponding key point storage nodes
Figure BDA0001742109880000076
The type of the node is a key point node, and the attribute of the node is a storage space required for storing a key point file
Figure BDA0001742109880000077
Then, it is calculated based on the following formula
Figure BDA0001742109880000081
And
Figure BDA0001742109880000082
data transmission rate of (2):
Figure BDA0001742109880000083
and adding a connection
Figure BDA0001742109880000084
And
Figure BDA0001742109880000085
is not limited by
Figure BDA0001742109880000086
The type of the edge is a storage edge, and the attribute of the edge is
Figure BDA0001742109880000087
And
Figure BDA0001742109880000088
and will be at the data transmission rate of
Figure BDA0001742109880000089
Adding into V, adding
Figure BDA00017421098800000810
Added to the set E.
c 2: traverse the redundancy strategy List RD [ 2 ]]For currently traversed sub-service siCorresponding main node
Figure BDA00017421098800000811
Adding corresponding redundant nodes
Figure BDA00017421098800000812
And, for any master node in set V
Figure BDA00017421098800000813
If it is with
Figure BDA00017421098800000814
Directly connected with each other, and added with an edge
Figure BDA00017421098800000815
The edge attribute is
Figure BDA00017421098800000816
And
Figure BDA00017421098800000829
the type of edge is a redundant edge, will
Figure BDA00017421098800000817
Added to the set E.
Step d: sequencing connection edges between main nodes in the graph model according to the descending order of data transmission rate, sequentially mapping head and tail nodes of the edges to physical machines of the cloud data center, distributing virtual machines on the physical machines according to requirements, and deploying sub-services corresponding to the nodes to the virtual machines. A preferred implementation is:
d 1: all the main edges in the set E are added to the list Ranked _ EM [ ], and the list Ranked _ EM [ ] is sorted from high to low in data transmission rate.
d2 traversing the list Ranked _ EM [ 2 ]]For the currently traversed edge eiObtaining eiHead (e) ofi) And tail node tail (e)i)。
d3 if head (e)i) And tail (e)i) If the corresponding sub-service is not deployed, searching the sub-network of the data center, and if two physical machines exist
Figure BDA00017421098800000818
And
Figure BDA00017421098800000819
of a sub-network of
Figure BDA00017421098800000820
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure BDA00017421098800000821
is greater than tail (e)i) The resource requirement of (1) is
Figure BDA00017421098800000822
Push head (e)i) Resource demand of (c) allocates virtual machines, head (e)i) The corresponding sub-service is deployed on the virtual machine
Figure BDA00017421098800000823
Push on tail (e)i) Allocate tail (e) to the virtual machinei) And d, deploying the corresponding sub-service to the virtual machine, and returning to the step d 2.
d4 if no subnet satisfying the condition is found in step d3, searching the cluster of the data center if there are two physical machines
Figure BDA00017421098800000824
And
Figure BDA00017421098800000825
is satisfied by
Figure BDA00017421098800000826
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure BDA00017421098800000827
is greater than tail (e)i) The resource requirement of (1) is
Figure BDA00017421098800000828
Push head (e)i) Resource demand of (c) allocates virtual machines, head (e)i) The corresponding sub-service is deployed on the virtual machine
Figure BDA0001742109880000091
Push on tail (e)i) Allocate tail (e) to the virtual machinei) And d, deploying the corresponding sub-service to the virtual machine, and returning to the step d 2.
d5 if there are no clusters satisfying the condition at step d4, then two physical machines are sought from the entire data center
Figure BDA0001742109880000092
And
Figure BDA0001742109880000093
satisfy the requirement of
Figure BDA0001742109880000094
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure BDA0001742109880000095
is greater than tail (e)i) Resource requirement of (c), head (e)i) Corresponding sub-service deployment to
Figure BDA0001742109880000096
tail(ei) Corresponding sub-service deployment to
Figure BDA0001742109880000097
Return to step d 2.
d6 if there is only head (e)i) Not yet deployed to deploy tail (e)i) Is the center of the physical machine, and searches for the residual resource from near to far to be larger than head (e)i) The physical machine of resource demand of (1), on which head (e) is pressedi) Allocating virtual machines to the demand of resources, head (e)i) The corresponding sub-service is deployed on the virtual machine, and the step d2 is returned.
d7 if there is only tail (e)i) Not yet deployed to deploy head (e)i) Is the center of the physical machine, and searches for the residual resource larger than tail (e) from near to fari) The physical machine on which tail (e) is pressedi) Allocating virtual machines to the demand for resources, tail (e)i) The corresponding sub-service is deployed on the virtual machine, and the step d2 is returned.
Step e: and sequencing edges between the main nodes and the key point storage nodes in the graph model according to the descending order of the data transmission rate, sequentially mapping tail nodes of the edges to a physical machine of the cloud data center, and distributing storage space for the key points on the physical machine. A preferred implementation is:
e 1: all the memory edges in the set E are added into a list Ranked _ EC [ ], and the Ranked _ EC [ ] is sorted from high to low according to the data transmission rate.
e2 traversing Ranked _ EC 2]For the currently traversed edge eiObtaining head node head (e)i) And tail node tail (e)i)。
e3 with head (e)i) The deployed physical machine is used as a center, and the residual resources are searched from near to far and are larger than tail (e)i) The physical machine of resource demand of (1), the physical machine being allocated as head (e)i) The key point of (2) stores the physical machine.
Step f: and sequencing the rest edges in the graph model according to the descending order of the data transmission rate, sequentially mapping the tail nodes to a physical machine of the cloud data center, and distributing virtual machines for the redundant service nodes on the physical machine. A preferred implementation is:
f 1: and adding all redundant edges in the set E into a list Ranked _ ER [ ], and sorting the Ranked _ ER [ ] from high to low according to the data transmission rate.
f2 traversal of Ranked _ ER [ alpha ]]For the currently traversed edge eiObtaining head node head (e)i) And tail node tail (e)i)。
f3 with head (e)i) The deployed physical machine is used as a center, and the residual resources are searched from near to far and are larger than tail (e)i) Physical machine of resource demand on which tail (e) is pressedi) Allocating virtual machines to the demand for resources, tail (e)i) A corresponding sub-service is deployed on the virtual machine.
The above description is only exemplary of the present application and should not be taken as limiting the present application, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present application should be included in the scope of protection of the present application.

Claims (7)

1. A service deployment method with enhanced reliability in a cloud environment is characterized by comprising the following steps:
a: modeling the complex service into a graph model, wherein a main node of the graph represents sub-services of the complex service, and edges of the graph represent data transmission rates among the sub-services;
b: traversing each sub-service of the complex service, and selecting key point storage or resource redundancy for each sub-service of the complex service as a reliability strategy of the sub-service according to the characteristics of the sub-service;
c: expanding the graph model based on the selected reliability strategy, and respectively adding a key point storage node or a redundant node for each main node;
d: sequencing all edges connected with the main nodes in the graph model according to the descending order of the data transmission rate, sequentially mapping head and tail nodes of the edges to a physical machine of the cloud data center, distributing virtual machines on the physical machine according to requirements, and deploying sub-services corresponding to the nodes to the virtual machines;
e: sequencing all edges connecting the main nodes and the key point storage nodes in the graph model according to the descending order of the data transmission rate, sequentially mapping tail nodes of the edges to a physical machine of the cloud data center, and distributing storage space for the key points on the physical machine;
f: and sequencing the rest edges in the graph model according to the descending order of the data transmission rate, sequentially mapping the tail nodes to the physical machine of the cloud data center, and distributing the virtual machines for the redundant service nodes on the physical machine.
2. The method of claim 1, wherein b comprises:
traversing all sub-services of the complex service, for a currently traversed sub-service siAccording to the time T _ exe(s) required for uninterrupted executioni) Time T _ cp(s) required for one key point storage for the sub-servicei) And an acceptable period T _ interval(s) for storing the key pointsi) The percentage of dead time perc is calculated based on the following formula:
Figure FDA0002465047320000011
then, comparing perc with an acceptability threshold delta, and if perc is greater than delta, selecting resource redundancy for the sub-service as a reliability strategy thereof; if perc is less than or equal to delta, the key point storage is selected for the sub-service as its reliability policy.
3. The method according to claim 1 or 2, wherein said a comprises:
a 1: constructing an empty graph model G (V, E) for the complex service, wherein the node set V and the edge set E are empty;
a 2: for complex services S ═ S1,s2...si...smax(s)Each sub-service s ofiAdding a graph node
Figure FDA0002465047320000021
The type of the node is a main node, the attribute of the node comprises the requirements of the sub-service on CPU resources, memory resources and disk resources, and the node is added into a node set V of G;
a 3: for sub-services s with data interactioniAnd sjIs its corresponding node
Figure FDA0002465047320000022
And
Figure FDA0002465047320000023
adding a side
Figure FDA0002465047320000024
And adding the edge to an edge set E of G, wherein the type of the edge is a main edge, and the attribute of the edge is a sub-service siAnd sjThe data transfer rate between.
4. The method of claim 1 or 2, wherein c comprises:
c 1: traversing and selecting key points to store as the sub-service of the reliability strategy, and performing the current traversal on the sub-service siCorresponding main node
Figure FDA0002465047320000025
Adding corresponding key point storage nodes
Figure FDA0002465047320000026
The type of the node is a key point node, and the attribute of the node is the storage required by storing the key point fileStorage space
Figure FDA0002465047320000027
Then, it is calculated based on the following formula
Figure FDA0002465047320000028
And
Figure FDA0002465047320000029
data transmission rate of (2):
Figure FDA00024650473200000210
and adding a connection
Figure FDA00024650473200000211
And
Figure FDA00024650473200000212
is not limited by
Figure FDA00024650473200000213
The type of the edge is a storage edge, and the attribute of the edge is
Figure FDA00024650473200000214
And
Figure FDA00024650473200000215
and will be at the data transmission rate of
Figure FDA00024650473200000216
Added to the node set V of the graph model, will
Figure FDA00024650473200000217
Adding the edge set E into the graph model;
c 2: traversing and selecting resource redundancy as a sub-service of the reliability strategy, and thenPre-traversed sub-services siCorresponding main node
Figure FDA00024650473200000218
Adding corresponding redundant nodes
Figure FDA00024650473200000219
And, for any master node in set V
Figure FDA00024650473200000220
If it is with
Figure FDA00024650473200000221
Directly connected with each other, and added with an edge
Figure FDA00024650473200000222
The edge attribute is
Figure FDA00024650473200000223
And
Figure FDA00024650473200000224
the type of edge is a redundant edge, will
Figure FDA0002465047320000031
Added to the set E.
5. The method of claim 1 or 2, wherein d comprises:
d 1: adding all main edges in an edge set E of the graph model into a list Ranked _ EM [ ], and sorting the list Ranked _ EM [ ] from high to low according to the data transmission rate;
d2 traversing the list Ranked _ EM [ 2 ]]For the currently traversed edge eiObtaining eiHead (e) ofi) And tail node tail (e)i)；
d3 if head (e)i) And tail (e)i) Corresponding sub-service has not yet been completedDeploying, searching the subnet of the data center, if two physical machines exist
Figure FDA0002465047320000032
And
Figure FDA0002465047320000033
of a sub-network of
Figure FDA0002465047320000034
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure FDA0002465047320000035
is greater than tail (e)i) The resource requirement of (1) is
Figure FDA0002465047320000036
Push head (e)i) Resource demand of (c) allocates virtual machines, head (e)i) The corresponding sub-service is deployed on the virtual machine
Figure FDA0002465047320000037
Push on tail (e)i) Allocate tail (e) to the virtual machinei) Deploying the corresponding sub-service to the virtual machine, and returning to the step d 2;
d4 if no subnet satisfying the condition is found in step d3, searching the cluster of the data center if there are two physical machines
Figure FDA0002465047320000038
And
Figure FDA0002465047320000039
is satisfied by
Figure FDA00024650473200000310
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure FDA00024650473200000311
is greater than tail (e)i) The resource requirement of (1) is
Figure FDA00024650473200000312
Push head (e)i) Resource demand of (c) allocates virtual machines, head (e)i) The corresponding sub-service is deployed on the virtual machine
Figure FDA00024650473200000313
Push on tail (e)i) Allocate tail (e) to the virtual machinei) Deploying the corresponding sub-service to the virtual machine, and returning to the step d 2;
d5 if there are no clusters satisfying the condition at step d4, then two physical machines are sought from the entire data center
Figure FDA00024650473200000314
And
Figure FDA00024650473200000315
satisfy the requirement of
Figure FDA00024650473200000316
Is greater than head (e)i) The resource requirements of (a) are determined,
Figure FDA00024650473200000317
is greater than tail (e)i) Resource requirement of (c), head (e)i) Corresponding sub-service deployment to
Figure FDA00024650473200000318
tail(ei) Corresponding sub-service deployment to
Figure FDA00024650473200000319
Returning to step d 2;
d6 if there is only head (e)i) Not yet deployed to deploy tail(ei) Is the center of the physical machine, and searches for the residual resource from near to far to be larger than head (e)i) The physical machine of resource demand of (1), on which head (e) is pressedi) Allocating virtual machines to the demand of resources, head (e)i) Deploying the corresponding sub-service on the virtual machine, and returning to the step d 2;
d7 if there is only tail (e)i) Not yet deployed to deploy head (e)i) Is the center of the physical machine, and searches for the residual resource larger than tail (e) from near to fari) The physical machine on which tail (e) is pressedi) Allocating virtual machines to the demand for resources, tail (e)i) The corresponding sub-service is deployed on the virtual machine, and the step d2 is returned.
6. The method of claim 4, wherein e comprises:
e 1: adding all storage edges in an edge set E of the graph model into a list Ranked _ EC [ ], and sorting the Ranked _ EC [ ] from high to low according to the data transmission rate;
e2 traversing Ranked _ EC 2]For the currently traversed edge eiObtaining head node head (e)i) And tail node tail (e)i)；
e3 with head (e)i) The deployed physical machine is used as a center, and the residual resources are searched from near to far and are larger than tail (e)i) The physical machine of resource demand of (1), the physical machine being allocated as head (e)i) The key point of (2) stores the physical machine.
7. The method of claim 4, wherein f comprises:
f 1: adding all redundant edges in an edge set E of the graph model into a list Ranked _ ER [ ], and sorting the Ranked _ ER [ ] from high to low according to the data transmission rate;
f2 traversal of Ranked _ ER [ alpha ]]For the currently traversed edge eiObtaining head node head (e)i) And tail node tail (e)i)；
f3 with head (e)i) The deployed physical machine is used as a center, and the residual resources are searched from near to far and are larger than tail (e)i) Physical machine of resource demand on which tail (e) is pressedi) Allocating virtual machines to the demand for resources, tail (e)i) A corresponding sub-service is deployed on the virtual machine.
CN201810824345.7A 2018-07-25 2018-07-25 Service deployment method for enhancing reliability in cloud environment Active CN109040214B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810824345.7A CN109040214B (en) 2018-07-25 2018-07-25 Service deployment method for enhancing reliability in cloud environment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810824345.7A CN109040214B (en) 2018-07-25 2018-07-25 Service deployment method for enhancing reliability in cloud environment

Publications (2)

Publication Number Publication Date
CN109040214A CN109040214A (en) 2018-12-18
CN109040214B true CN109040214B (en) 2020-07-17

Family

ID=64645027

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810824345.7A Active CN109040214B (en) 2018-07-25 2018-07-25 Service deployment method for enhancing reliability in cloud environment

Country Status (1)

Country Link
CN (1) CN109040214B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109526012A (en) * 2019-01-24 2019-03-26 重庆邮电大学 A kind of LoRaWAN network spreading factor distribution method based on reliability

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102768670A (en) * 2012-05-31 2012-11-07 哈尔滨工程大学 Webpage clustering method based on node property label propagation
CN104579850A (en) * 2015-01-27 2015-04-29 北京邮电大学 Quality of service (QoS) prediction method for Web service under mobile Internet environment
CN104809168A (en) * 2015-04-06 2015-07-29 华中科技大学 Partitioning and parallel distribution processing method of super-large scale RDF graph data
CN106708978A (en) * 2016-12-07 2017-05-24 北京邮电大学 Bipartite graph-based service recommendation method and apparatus

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10356206B2 (en) * 2015-11-08 2019-07-16 Vmware, Inc. Deploying an application in multiple cloud computing environments

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102768670A (en) * 2012-05-31 2012-11-07 哈尔滨工程大学 Webpage clustering method based on node property label propagation
CN104579850A (en) * 2015-01-27 2015-04-29 北京邮电大学 Quality of service (QoS) prediction method for Web service under mobile Internet environment
CN104809168A (en) * 2015-04-06 2015-07-29 华中科技大学 Partitioning and parallel distribution processing method of super-large scale RDF graph data
CN106708978A (en) * 2016-12-07 2017-05-24 北京邮电大学 Bipartite graph-based service recommendation method and apparatus

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于时空图的移动对象聚集模式挖掘方法;张峻铭，李静林，王尚广,刘志晗,袁泉,杨放春;《软件学报》;20151103;全文 *

Also Published As

Publication number Publication date
CN109040214A (en) 2018-12-18

Similar Documents

Publication Publication Date Title
US11704144B2 (en) Creating virtual machine groups based on request
US10496627B2 (en) Consistent ring namespaces facilitating data storage and organization in network infrastructures
WO2016090946A1 (en) Method and device for mapping virtual data centre resources
US9888065B2 (en) Defragmenting clusters with reserved resources
CN103929454B (en) The method and system of load balancing storage in a kind of cloud computing platform
CN111182067B (en) Data writing method and device based on interplanetary file system IPFS
JP2012524946A5 (en)
CN104679594B (en) A kind of middleware distributed computing method
CN111177017B (en) Memory allocation method and device
CN105357322A (en) Virtual machine distribution method based on topology partition
CN108282522A (en) Data storage access method based on dynamic routing and system
CN107070645A (en) Compare the method and system of the data of tables of data
CN105357042A (en) High-availability cluster system, master node and slave node
CN112948279A (en) Method, apparatus and program product for managing access requests in a storage system
CN109040214B (en) Service deployment method for enhancing reliability in cloud environment
CN106293509A (en) Date storage method and system
CN109298949B (en) Resource scheduling system of distributed file system
CN113835823A (en) Resource scheduling method and device, electronic equipment and computer readable storage medium
CN108604231A (en) Mirror processing method and computing device
CN113055448B (en) Metadata management method and device
CN112398668B (en) IaaS cluster-based cloud platform and node switching method
CN114157583A (en) Reliability-based network resource heuristic mapping method and system
CN109787899B (en) Data partition routing method, device and system
CN117041258B (en) BIM engineering information management system
WO2016188135A1 (en) Cpu resource configuration method for cluster router and cluster router

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