CN109828829B - Virtual machine rapid evacuation method based on disaster early warning time - Google Patents
Virtual machine rapid evacuation method based on disaster early warning time Download PDFInfo
- Publication number
- CN109828829B CN109828829B CN201910087902.6A CN201910087902A CN109828829B CN 109828829 B CN109828829 B CN 109828829B CN 201910087902 A CN201910087902 A CN 201910087902A CN 109828829 B CN109828829 B CN 109828829B
- Authority
- CN
- China
- Prior art keywords
- evacuation
- virtual machine
- bandwidth
- virtual
- early warning
- 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
Images
Abstract
The invention discloses a method for quickly evacuating virtual machines based on disaster early warning time, and belongs to the technical field of cloud computing. Aiming at the serious threat of a large-scale disaster to physical nodes of a backbone network, the method reconstructs and evacuates the threatened virtual network and related nodes thereof according to disaster early warning time, the memory size of the virtual machine and the downtime of the virtual machine. The method adopts a post-copy migration technology, allocates the optimal evacuation path and initial bandwidth for the virtual machine to be evacuated under the migration flow balance principle, dynamically adjusts the bandwidth on the evacuation path according to the network resource condition, and maximizes the evacuation completion rate of the virtual machine within the early warning time.
Description
Technical Field
The invention belongs to the technical field of cloud computing, and particularly relates to a rapid virtual machine evacuation method based on disaster early warning time.
Background
The cloud computing virtualizes underlying physical network resources through a virtualization technology and puts the underlying physical network resources into a shared resource pool, so that flexible and rentable resources are provided for users. The user can rent proper resources according to the requirement of the user to build the needed virtual network. Network virtualization may map multiple virtual networks to the same underlying physical network, with virtual nodes typically existing in the form of virtual machines, and with multiple virtual machines typically existing in a server that is an underlying physical network node. Virtual machines are interconnected by virtual links, while servers or data centers are interconnected by underlying physical links.
The virtual machine migration has extremely important significance for network load balancing, network energy saving, survivability problems and the like. The current online migration technology of virtual machines mainly comprises: the method comprises a memory-based pre-copy migration technology and a memory-based post-copy migration technology. Compared with pre-copy migration, post-copy migration omits the iterative transmission process of the dirty codes, so that the migrated data volume is less and the migration time is shorter.
Large-scale disaster events, such as: earthquake can cause serious threat and damage to the underlying physical network, and further affect the virtual network mapped on the earthquake. The early warning time of the earthquake is mainly determined by the propagation velocity of the earthquake wave and generally ranges from a few seconds to tens of seconds. How to rapidly evacuate a virtual machine within a short early warning time, reduce the influence of disasters on virtual network services and reduce the loss of service providers and users becomes a problem which needs to be solved urgently in the field of cloud computing.
The invention comprehensively considers disaster early warning time, the memory size of the virtual machine and the downtime of the virtual machine, adopts the post-copy migration technology and establishes the optimal evacuation strategy for the threatened virtual machine, thereby maximizing the evacuation completion rate of the virtual machine within the early warning time.
Disclosure of Invention
The present invention is directed to solving the above problems of the prior art. The virtual machine rapid evacuation method based on the disaster early warning time maximizes the evacuation completion rate of the virtual machines within the early warning time. The technical scheme of the invention is as follows:
a method for quickly evacuating virtual machines based on disaster early warning time is characterized in that a post-copy migration technology is adopted to reconstruct and evacuate threatened virtual networks and related nodes thereof according to the disaster early warning time, the memory size of the virtual machines and the downtime of the virtual machines, and specifically comprises the following steps:
101. let the current time t c =0, all virtual machines to be evacuated { m i Amount of evacuation data
Evacuation end time
T is early warning time;
102. if virtual machine set is to be evacuated
For each virtual machine M to be evacuated in M i Reconstructing the virtual network, jumping to step 103, otherwise, jumping to step 104;
103. virtual machine m to be evacuated for completion of reconstruction i Calculating and allocating egress routes and bandwidth
And the virtual machine m to be evacuated with the evacuation route and the bandwidth successfully allocated i Moving into a set of virtual machines M performing an evacuation c ；
104. If it is not
To M c All m in i Evacuation is executed, and the step 105 is skipped, otherwise, the algorithm is ended;
105. to the downtime ending time
M of (a) i (∈M c ) Upgrade evacuation Bandwidth
To the upper limit of the bandwidth capacity of the evacuation path, and updating the evacuation ending time
106. In that
During, if m is present i (∈M c ) Ending the downtime to
Will correspond to m i Bandwidth of evacuation
Upgrade to the upper limit of the bandwidth capacity of the evacuation path and update the corresponding evacuation ending time
Jumping to step 106 to continue execution, otherwise, jumping to step 107;
107. when in use
Then m for completing evacuation at the current moment is marked i And updating the network resources to connect m i From the set M c And (4) deleting, and jumping to step 102.
Further, said step 102 is to evacuate each M in the set M of virtual machines to be evacuated i The step of reconfiguring the virtual network comprises:
1) Outside the risk area, for the virtual machine m to be evacuated i Finding candidate target nodes with sufficient node resources, so that A i Each node in the network has a distance of no more than h hops from the target node and all the nodes are putGo into set S i In which A is i For virtual machine m i The physical node set mapped by the adjacent virtual machine;
2) If it is not
To S i In each alternative target node, the Dijkstra minimum cost routing algorithm is adopted to calculate the routing and bandwidth resources required by the reconstruction of the corresponding virtual network, and the target nodes which can not find the reconstruction resources are collected from a set S i Deleting, otherwise, dividing m i Delete from set M;
3) If it is used
And (3) selecting the node with the minimum reconstruction resource cost in the step (2) as a target node, and reserving corresponding nodes and bandwidth resources in the network.
Further, step 103 is to complete reconstructed m i Allocating bandwidth
The evacuation bandwidth calculation formula used is defined as:
in the formula (1), D i Representing a virtual machine m i T represents the early warning time, tau g And representing the upper limit of the downtime of the virtual machine.
Further, step 103 is to complete reconstructed m i The evacuation route is calculated by using Dijkstra algorithm, and a link cost calculation formula is defined as follows:
in the formula (2), c l Is the cost of the physical link l (E E), E is the set of physical network links, VFor a set of physical network nodes, | V | represents the number of physical network nodes, n represents the number of evacuation paths through link l, first part of equation (2)
Number of physical link hops for restricting evacuation route, second part
For equalizing the traffic distribution of the evacuation paths.
Further, the steps 105 and 106 update the respective evacuation end time
Is defined as:
in the formula (3), the first and second groups,
representing a virtual machine m i Evacuation end time of t c Indicating the current time, D i Representing a virtual machine m i The amount of data stored in the memory of the memory,
representing a current virtual machine m i The amount of data that has been evacuated,
representing a current virtual machine m i The evacuation bandwidth of (1).
Further, the step 106 is at
During the period of time (T) when the user is in the normal state,
is defined as follows:
in the formula (4), the first and second groups of the chemical reaction are shown in the specification,
virtual machine m representing the next evacuation completed first i At the corresponding evacuation-end time point, the evacuation-end time point,
a set of virtual machine evacuation end times representing future virtual machine evacuation completions.
The invention has the following advantages and beneficial effects:
aiming at the serious influence of large-scale disaster events on the virtual network service, the invention provides a rapid virtual machine evacuation method based on disaster early warning time, which is used for reconstructing and evacuating the virtual network threatened by the disaster and related nodes thereof. According to the method, a post-copy migration technology is adopted, and a virtual machine evacuation bandwidth calculation formula is designed according to disaster early warning time, the size of a virtual machine memory and the downtime of the virtual machine, so that the initial bandwidth requirement of the virtual machine to be evacuated is guaranteed; designing a link cost formula according to a migration flow balance principle, and searching an optimal evacuation path for the virtual machine to be evacuated by adopting a Dijkstra algorithm; the evacuation bandwidth of the virtual machine is adaptively adjusted according to the network resource condition, and the utilization rate of the network bandwidth resource is dynamically optimized, so that the evacuation completion rate of the virtual machine within the early warning time is maximized.
Drawings
Fig. 1 is a flow chart of a virtual machine rapid evacuation method based on disaster warning time according to a preferred embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described in detail and clearly with reference to the accompanying drawings. The described embodiments are only some of the embodiments of the present invention.
The technical scheme for solving the technical problems is as follows:
the concepts and models to which the present disclosure relates are as follows:
1. physical network model
Assume a physical network topology of G (V, E), where V represents a set of physical nodes, each having data storage and routing capabilities, and E represents a set of physical links.
2. Other symbols referred to in the context of the present invention are described below:
m: set of virtual machines to be evacuated, m i ∈M
·A i : virtual machine m i Of the physical node set mapped by the adjacent virtual machine
·S i : virtual machine m i Remapped candidate target physical node set
T: early warning time threshold value
·τ g : upper limit of virtual machine downtime
·
Virtual machine m i End of downtime
·
Virtual machine m i Evacuation end time of
·D i : virtual machine m i Amount of memory data of
·t c : current time of day
·
Current virtual machine m i Bandwidth of evacuation
·
Current virtual machine m i Volume of evacuated data
·M c : set of virtual machines performing an evacuation
·
A set of virtual machine evacuation end times representing future virtual machine evacuation completions.
The technical scheme of the invention is explained as follows:
1. virtual network reconstruction method
Step 1: outside the risk area, m i Finding candidate target nodes with sufficient node resources, so that A i Each node in the set has a distance of no more than h hops from the target node, and all the nodes are put into a set S i Performing the following steps;
step 2: if it is not
To S i In each target node, calculating the route and bandwidth resources required by reconstructing the corresponding virtual network, and selecting the candidate target nodes from the S i Deleting, otherwise, dividing m i Delete from set M;
and step 3: if it is not
And selecting the candidate node with the minimum reconstruction resource cost as the target node, and reserving corresponding nodes and bandwidth resources in the network.
2. Evacuation bandwidth calculation formula
Virtual machine m i Evacuation bandwidth
Is defined as:
in the formula (1), D i Representing a virtual machine m i T represents the early warning time, tau g And representing the upper limit of the downtime of the virtual machine.
3. Evacuation route link cost formula
VirtualizationMachine m i The calculation of the evacuation route adopts Dijkstra algorithm, wherein a link cost formula is defined as:
in the formula (2), c l For the cost of link l (∈ E), | V | represents the number of physical network nodes, and n represents the number of egress paths through link l. First part of equation (2)
Number of physical link hops for restricting evacuation route, second part
For equalizing the traffic distribution of the evacuation paths.
4. Evacuation end time
Is defined as:
in the formula (3), the first and second groups,
representing a virtual machine m i Evacuation end time of t c Indicating the current time, D i Representing a virtual machine m i The amount of data stored in the memory of the memory,
representing a current virtual machine m i The amount of data that has been evacuated,
representing a current virtual machine m i The evacuation bandwidth of (1).
A rapid virtual machine evacuation method based on disaster early warning time is characterized by comprising the following steps:
step 1: initialization t c ＝0，
Step 2: if it is not
For each M in M i Reconstructing the virtual network, and jumping to the step 3, otherwise, jumping to the step 4;
and step 3: m for complete reconstruction according to equation (1) i Calculating and allocating egress routes and bandwidth
And allocates the evacuation route and bandwidth successfully to m i Move into set M c ；
And 4, step 4: if it is not
To M c All m in i Performing evacuation, and jumping to the step 5, otherwise, ending the algorithm;
and 5: to pair
M of (a) i (∈M c ) Upgrading evacuation bandwidth
To the upper limit of the bandwidth capacity of the evacuation path and updating according to the formula (3)
Step 6: in that
During, if m is i (∈M c ) Ending the downtime to
Will correspond to m i Bandwidth of evacuation
Upgrading to the upper limit of the bandwidth capacity of the evacuation path, and updating according to the formula (3)
Jumping to the step 6, otherwise, jumping to the step 7;
and 7: when the temperature is higher than the set temperature
Then m for completing evacuation at the current moment is marked i And updating the network resources to connect m i From the set M c And (5) deleting, and jumping to the step 2.
The above examples are to be construed as merely illustrative and not limitative of the remainder of the disclosure. After reading the description of the invention, the skilled person can make various changes or modifications to the invention, and these equivalent changes and modifications also fall into the scope of the invention defined by the claims.
Claims (5)
1. A method for quickly evacuating virtual machines based on disaster early warning time is characterized in that a post-copy migration technology is adopted to reconstruct and evacuate threatened virtual networks and related nodes thereof according to the disaster early warning time, the memory size of the virtual machines and the downtime of the virtual machines, and specifically comprises the following steps:
101. let the current time t c =0, all virtual machines to be evacuated { m i Amount of evacuation data
Evacuation end time
T is early warning time;
102. if virtual machine set is to be evacuated
For each virtual machine M to be evacuated in M i Reconstructing the virtual network, jumping to step 103, otherwise, jumping to step 104;
103. virtual machines m to be evacuated for the purpose of completing a reconstruction i Calculating and allocating egress routes and bandwidth
And the virtual machine m to be evacuated with the evacuation route and the bandwidth successfully allocated i Moving into a set of virtual machines M performing an evacuation c ；
104. If it is not
To M c All m in i Evacuation is executed, and the step 105 is skipped, otherwise, the algorithm is ended;
105. to the downtime ending time
M of i (∈M c ) Upgrade evacuation Bandwidth
To the upper limit of the bandwidth capacity of the evacuation path, and updating the evacuation ending time
106. In that
During, if m is present i (∈M c ) Ending the downtime to order
Will correspond to m i Bandwidth of evacuation
Upgrade to the upper limit of the bandwidth capacity of the evacuation path and update the corresponding evacuation ending time
Jumping to step 106 to continue execution, otherwise, jumping to step 107;
107. when in use
Then m for completing evacuation at the current moment is marked i And updating network resources to get m i From the set M c Deleting, jumping to step 102;
said step 106 is at
During the period of time in which the air is being discharged,
is defined as follows:
in the formula (4), the first and second groups,
virtual machine m representing the next evacuation completed first i At the corresponding evacuation-end time point, the evacuation-end time point,
a set of virtual machine evacuation end times representing future virtual machine evacuation completions.
2. The method for virtual machine rapid evacuation based on disaster early warning time as claimed in claim 1, wherein said step 102 is to evacuate each M in virtual machine set M to be evacuated i The step of reconfiguring the virtual network comprises:
1) Outside the risk area, for the virtual machine m to be evacuated i Finding candidate target nodes with enough node resources, so that A i Each node in the set has a distance of no more than h hops from the target node, and all the nodes are put into a set S i In which A is i For virtual machine m i The physical node set mapped by the adjacent virtual machine;
2) If it is used
To S i In each alternative target node, the Dijkstra minimum cost routing algorithm is adopted to calculate the routing and bandwidth resources required by the reconstruction of the corresponding virtual network, and the target nodes which can not find the reconstruction resources are gathered from the set S i Deleting, otherwise, dividing m i Delete from set M;
3) If it is not
And (3) selecting the node with the minimum reconstruction resource cost in the step (2) as a target node, and reserving corresponding nodes and bandwidth resources in the network.
3. The method for virtual machine fast evacuation based on disaster warning time as claimed in claim 2, wherein said step 103 is m for completing reconstruction i Allocating bandwidth
The evacuation bandwidth calculation formula used is defined as:
in the formula (1), D i Representing a virtual machine m i T represents the early warning time, tau g And representing the upper limit of the downtime of the virtual machine.
4. The method for virtual machine fast evacuation based on disaster warning time as claimed in one of claims 1-3, wherein said step 103 is m for completing reconstruction i The evacuation route is calculated by adopting a Dijkstra algorithm, and a link cost calculation formula is defined as follows:
in the formula (2), c l Is the cost of a physical link l (epsilon E), E is a physical network link set, V is a physical network node set, | V | represents the number of physical network nodes, n represents the number of evacuation paths passing through the link l, and the first part of the formula (2)
Number of physical link hops for restricting evacuation route, second part
For equalizing the traffic distribution of the evacuation paths.
5. Method for virtual machine fast evacuation based on disaster warning time according to one of claims 1-3, characterized in that steps 105 and 106 update the corresponding evacuation end time
Is defined as:
in the formula (3), the first and second groups,
representing a virtual machine m i Evacuation end time of t c Indicating the current time, D i Representing a virtual machine m i The amount of data stored in the memory of the memory,
representing a current virtual machine m i The amount of data that has been evacuated,
representing a current virtual machine m i The evacuation bandwidth of (1).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910087902.6A CN109828829B (en) | 2019-01-22 | 2019-01-22 | Virtual machine rapid evacuation method based on disaster early warning time |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910087902.6A CN109828829B (en) | 2019-01-22 | 2019-01-22 | Virtual machine rapid evacuation method based on disaster early warning time |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109828829A CN109828829A (en) | 2019-05-31 |
| CN109828829B true CN109828829B (en) | 2022-10-18 |
Family
ID=66862996
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910087902.6A Active CN109828829B (en) | 2019-01-22 | 2019-01-22 | Virtual machine rapid evacuation method based on disaster early warning time |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109828829B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101609419A (en) * | 2009-06-29 | 2009-12-23 | 北京航空航天大学 | The data back up method and the device of the migration of virtual machine consistently online |
| CN103259589A (en) * | 2013-05-27 | 2013-08-21 | 重庆邮电大学 | Virtual backup capacity destroy-resistance protection method of energy-saving optical network |
| CN104010028A (en) * | 2014-05-04 | 2014-08-27 | 华南理工大学 | Dynamic virtual resource management strategy method for performance weighting under cloud platform |
| EP2879053A1 (en) * | 2013-09-30 | 2015-06-03 | Huawei Technologies Co., Ltd. | Virtual machine memory data migration method, related apparatus, and cluster system |
| CN108196935A (en) * | 2017-12-06 | 2018-06-22 | 南京邮电大学 | A kind of energy saving moving method of virtual machine towards cloud computing |
| CN108900358A (en) * | 2018-08-01 | 2018-11-27 | 重庆邮电大学 | Virtual network function dynamic migration method based on deepness belief network resource requirement prediction |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8370560B2 (en) * | 2009-11-16 | 2013-02-05 | International Business Machines Corporation | Symmetric live migration of virtual machines |
| US9558005B2 (en) * | 2014-05-19 | 2017-01-31 | Intel Corporation | Reliable and deterministic live migration of virtual machines |
-
2019
- 2019-01-22 CN CN201910087902.6A patent/CN109828829B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101609419A (en) * | 2009-06-29 | 2009-12-23 | 北京航空航天大学 | The data back up method and the device of the migration of virtual machine consistently online |
| CN103259589A (en) * | 2013-05-27 | 2013-08-21 | 重庆邮电大学 | Virtual backup capacity destroy-resistance protection method of energy-saving optical network |
| EP2879053A1 (en) * | 2013-09-30 | 2015-06-03 | Huawei Technologies Co., Ltd. | Virtual machine memory data migration method, related apparatus, and cluster system |
| CN104010028A (en) * | 2014-05-04 | 2014-08-27 | 华南理工大学 | Dynamic virtual resource management strategy method for performance weighting under cloud platform |
| CN108196935A (en) * | 2017-12-06 | 2018-06-22 | 南京邮电大学 | A kind of energy saving moving method of virtual machine towards cloud computing |
| CN108900358A (en) * | 2018-08-01 | 2018-11-27 | 重庆邮电大学 | Virtual network function dynamic migration method based on deepness belief network resource requirement prediction |
Non-Patent Citations (8)
| Title |
|---|
| "Disaster-prediction_based_virtual_network_mapping_against_multiple_regional_failures";Xiao Liu,et al;《IEEE》;20151231;全文 * |
| "Hierarchical Fairness Based Re-provisioning in Post-Disaster Telecom Networks ";Ning-Hai Bao, Ya-Kun Wu, Guo-Qing Su, Yuan Yuan, Da-Yong Luo;《IEEE》;20171231;全文 * |
| "Joint_Mapping_and_Routing_of_Virtual_Network_Functions_for_Improved_Disaster_Recovery_Support";D. Oliveira，et al;《IEEE》;20181231;全文 * |
| "Reliability Based Restoration in Post-Disaster Optical Networks";Ning-Hai Bao, Jing-Bo Chen, Yuan Yuan, Ming Kuang;《IEEE》;20171231;全文 * |
| "Reliability-sustainable_network_survivability_scheme_against_disaster_failures";Ning-Hai Bao, Guo-Qing Su, Ya-Kun Wu, Ming Kuang, Da-Yong Luo;《IEEE》;20171231;全文 * |
| "基于优化开销收益比的大规模虚拟网络映射问题的研究 ";李雄;《中国优秀硕士学位论文全文数据库》;20190115;全文 * |
| "基于链路生命期的光数据中心网络业务恢复方案";鲍宁海，袁园，刘自谦，匡明;《通信学报》;20180831;全文 * |
| 面向动态虚拟网络请求的虚拟网络映射算法;苑迎等;《计算机应用》;20170110(第01期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109828829A (en) | 2019-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111147307B (en) | Service function chain reliable deployment method based on deep reinforcement learning | |
| CN113708972B (en) | Service function chain deployment method and device, electronic equipment and storage medium | |
| US10660069B2 (en) | Resource allocation device and resource allocation method | |
| WO2017016220A1 (en) | Connectivity-aware virtual network embedding | |
| CN103812748B (en) | Mapping method of survivable virtual network | |
| CN107979409A (en) | Interorbital satellite communication method for routing and device | |
| CN103455363B (en) | Command processing method, device and physical host of virtual machine | |
| JP2017500816A (en) | Traffic engineering for large data center networks | |
| CN111327708B (en) | Optical network virtual mapping method based on topology perception | |
| Spinnewyn et al. | Resilient application placement for geo-distributed cloud networks | |
| US20190018698A1 (en) | Optimizing virtual machine placement for multi-destination traffic | |
| CN111698114B (en) | Determining and implementing optimized traffic plans with constrained optimization of egress peer-to-peer engineering | |
| CN113835823A (en) | Resource scheduling method and device, electronic equipment and computer readable storage medium | |
| US10873502B2 (en) | System and method for joint embedding and backup provisioning in virtual networks | |
| CN109828829B (en) | Virtual machine rapid evacuation method based on disaster early warning time | |
| CN111740853B (en) | Determining and implementing optimized business plans using egress peer-to-peer engineering | |
| Yao et al. | VM migration planning in software-defined data center networks | |
| CN105490959B (en) | Implementation method is embedded in based on the non-homogeneous bandwidth virtual data center that congestion is evaded | |
| JP6389811B2 (en) | Physical resource allocation device, physical resource allocation method, and program | |
| CN110058970B (en) | Double-virtual-machine quasi-synchronous evacuation method under disaster risk model | |
| CN115767325A (en) | Service function chain profit maximization mapping method and system | |
| CN113301583B (en) | Network slice remapping method, device and storage medium | |
| JP6586228B2 (en) | Route calculation control device and route calculation control method | |
| CN114466059B (en) | Method for providing reliable service function chain for mobile edge computing system | |
| CN115361284B (en) | Deployment adjustment method of virtual network function based on SDN |
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 |