CN116938943A - Cloud host scheduling method, device, equipment and storage medium - Google Patents
Cloud host scheduling method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN116938943A CN116938943A CN202311190639.6A CN202311190639A CN116938943A CN 116938943 A CN116938943 A CN 116938943A CN 202311190639 A CN202311190639 A CN 202311190639A CN 116938943 A CN116938943 A CN 116938943A
- Authority
- CN
- China
- Prior art keywords
- cloud host
- node
- affinity
- policy
- cloud
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 92
- 238000012163 sequencing technique Methods 0.000 claims abstract description 45
- 238000001914 filtration Methods 0.000 claims abstract description 14
- 230000005012 migration Effects 0.000 claims description 35
- 238000013508 migration Methods 0.000 claims description 35
- 238000004590 computer program Methods 0.000 claims description 13
- 230000004044 response Effects 0.000 claims description 6
- 230000008569 process Effects 0.000 abstract description 26
- 238000004891 communication Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000007726 management method Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- MWRWFPQBGSZWNV-UHFFFAOYSA-N Dinitrosopentamethylenetetramine Chemical compound C1N2CN(N=O)CN1CN(N=O)C2 MWRWFPQBGSZWNV-UHFFFAOYSA-N 0.000 description 1
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000003818 cinder Substances 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000010348 incorporation Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 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
- 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
-
- 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
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
-
- 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)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of computers, and provides a cloud host scheduling method, a device, equipment and a storage medium, wherein the method comprises the following steps: receiving a first request for scheduling the cloud hosts binding the affinity group; responding to the first request, and judging whether the first node meets a preset first strategy; the first strategy is an affinity group strategy determined based on the cloud host; under the condition that the first node does not meet the first strategy, filtering a plurality of second nodes based on the performance parameter information to obtain at least one alternative node; obtaining a sequencing result according to at least one alternative node, weight information corresponding to each alternative node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host; determining a target node to be migrated according to the sequencing result; and migrating the cloud host to the target node before the cloud host is started. The scheduling process of the invention is more flexible and is more suitable for actual use scenes.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a cloud host scheduling method, device, equipment, and storage medium.
Background
OpenStack cloud host scheduling refers to the process of distributing cloud host instances onto appropriate physical hosts in an OpenStack cloud computing platform.
In the prior art, in the process of dispatching the cloud host of the affinity group, the binding of the affinity group can only be carried out when the cloud host is created, and then the affinity group of the cloud host cannot be bound or unbound, so that the dispatching process of the cloud host is not flexible enough, and the adaptability of the existing cloud host dispatching method is poor.
Disclosure of Invention
The invention provides a cloud host scheduling method, a device, equipment and a storage medium, which are used for solving the defects of inflexible cloud host scheduling process and poor adaptability of the existing cloud host scheduling method, realizing the improvement of flexibility in cloud host scheduling and being more suitable for actual scenes.
In a first aspect, the present invention provides a cloud host scheduling method, including:
receiving a first request for scheduling the cloud hosts binding the affinity group;
responding to the first request, and judging whether the first node meets a preset first strategy; the first policy is an affinity group policy determined based on the cloud host;
filtering a plurality of second nodes based on performance parameter information under the condition that the first node does not meet the first policy to obtain at least one alternative node;
Obtaining a sequencing result according to the weight information corresponding to each candidate node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host;
determining a target node to be migrated according to the sequencing result;
and before the cloud host is started, migrating the cloud host to the target node.
Optionally, the obtaining a sorting result according to the at least one candidate node, the weight information corresponding to each candidate node, and a preset second policy includes:
judging whether the at least one alternative node meets the second strategy;
and under the condition that at least one candidate node meets the second strategy, carrying out weight ranking on the at least one candidate node by utilizing weight information corresponding to each at least one candidate node, and obtaining the ranking result.
Optionally, before the responding to the first request and determining whether the first node meets the preset first policy, the method further includes:
adding a distributed lock to the affinity group corresponding to the cloud host, wherein the distributed lock is used for indicating the migration serial execution of the cloud hosts in the same affinity group; the distributed lockset is provided with a unique KEY value KEY; and the affinity group corresponding to the cloud host has a unique identification ID.
Optionally, after the cloud host is migrated to the target node when the cloud host is powered on, the method further includes:
releasing the distributed lock.
Optionally, the determining the target node to be migrated according to the sorting result includes:
determining an alternative node corresponding to the maximum weight value in the sequencing result as the target node; or alternatively, the first and second heat exchangers may be,
according to the number K of the preset target nodes and the sequencing result, K candidate nodes with the maximum weight value are sequentially determined as the target nodes in the at least one candidate node; k is a positive integer.
Optionally, the migration is cold migration.
Optionally, the method further comprises:
and performing target operation on the cloud host and the affinity group at any moment when the state of the cloud host is in a shutdown state, wherein the target operation comprises at least one of binding the cloud host and the affinity group and unbinding the cloud host and the affinity group.
In a second aspect, the present invention provides a cloud host scheduling apparatus, including:
the receiving module is used for receiving a first request for scheduling the cloud host binding the affinity group;
the scheduling module is used for responding to the first request and judging whether the first node meets a preset first strategy or not; the first policy is an affinity group policy determined based on the cloud host; filtering a plurality of second nodes based on performance parameter information under the condition that the first node does not meet the first policy to obtain at least one alternative node; obtaining a sequencing result according to the weight information corresponding to each candidate node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host; determining a target node to be migrated according to the sequencing result;
And the migration module migrates the cloud host to the target node before the cloud host is started.
In a third aspect, the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements any one of the cloud host scheduling methods described above when executing the program.
In a fourth aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a cloud host scheduling method as described in any of the above.
In a fifth aspect, the present invention also provides a computer program product comprising a computer program which, when executed by a processor, implements a cloud host scheduling method as described in any of the above.
According to the cloud host scheduling method, the cloud host scheduling device, the cloud host scheduling equipment and the cloud host storage medium, whether a first node meets a preset first strategy is judged by receiving a first request for scheduling the cloud hosts binding the affinity group, wherein the first strategy is an affinity group strategy determined based on the cloud hosts; further, under the condition that the first node does not meet the first policy, filtering the plurality of second nodes based on the performance parameter information to obtain at least one alternative node; then, according to at least one alternative node, weight information corresponding to each alternative node and a preset second strategy, a sequencing result is obtained, wherein the second strategy comprises at least one of a weak affinity strategy determined based on a cloud host and a weak anti-affinity strategy determined based on the cloud host; determining a target node to be migrated according to the sequencing result; and finally, migrating the cloud host to the target node before the cloud host is started. According to the method, when the first node does not meet the preset first strategy, the plurality of second nodes are filtered based on the performance parameter information to obtain at least one alternative node, and then, according to the at least one alternative node, the weight information corresponding to each alternative node and the preset second strategy, a sequencing result is obtained, the target node to be migrated is determined according to the sequencing result, finally, the cloud host is migrated to the target node before the cloud host is started, and the cloud host scheduling process based on the first strategy and the second strategy is put before the cloud host is started, so that flexibility in cloud host scheduling is improved, and the method is more suitable for actual use scenes.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a cloud host scheduling method provided by the invention;
FIG. 2 is a second flow chart of the cloud host scheduling method according to the present invention;
FIG. 3 is a third flow chart of the cloud host scheduling method according to the present invention;
fig. 4 is a schematic structural diagram of a cloud host scheduling device provided by the invention;
fig. 5 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are 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 invention without making any inventive effort, are intended to be within the scope of the invention.
The terms first, second and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances such that the embodiments of the application are capable of operation in sequences other than those illustrated or otherwise described herein, and that the "first" and "second" distinguishing between objects generally are not limited to the number of objects, for example, the first node may be one or more. Furthermore, in the description and claims, "and/or" means at least one of the connected objects, and the character "/" generally means a relationship in which the associated object is an "or" before and after.
First, technical terms related to the embodiments of the present application will be described.
Cloud platform: the user's investment in infrastructure is converted into operating costs (Operating Expense, OPEX), and the user no longer has to plan for his own data center, nor has to expend effort on the internet technology IT management independent of his own camping service. They need only send an indication to the "cloud" to get different degrees and different types of information services.
OpenStack: is an open-source cloud computing platform, providing a series of components and tools for building and managing private and public cloud environments. The main components of OpenStack include a compute (Nova) component, a network (Neutron) component, a store (Cinder, swift) component, etc., which can work cooperatively to provide a complete cloud computing infrastructure service. The OpenStack has the advantages that the OpenStack is open-source, has high flexibility and expandability, and can be customized and configured according to actual requirements. In addition, openStack provides rich application program interface APIs and Command Line Interface (CLI) tools, which facilitate management and operation for users.
Distributed lock: a distributed lock is a mechanism for coordinating access to shared resources between processes in a distributed system. In a distributed environment, multiple processes may access the same shared resource at the same time, and in order to avoid creating a race condition, a distributed lock needs to be used to ensure that only one process can access the resource at the same time. There are various ways to implement the distributed lock, such as database-based, cache redis-based, zooKeeper-based, etc.
Affinity: cloud host affinity refers to a technique of deploying multiple instances of the same application on the same physical node or the same physical server cluster in order to improve the reliability and performance of the application in a cloud computing environment.
Node affinity: meaning that cloud hosts in the same affinity group must coexist on the same physical node;
node anti-affinity: meaning that cloud hosts in the same anti-affinity group must be distributed on different physical nodes;
weak affinity of nodes: the cloud hosts in the same weak affinity group preferably coexist on the same physical node, and when the resources of the physical node are insufficient, the strategy is not followed any more, and the cloud hosts can be distributed on other physical nodes;
weak anti-affinity of the nodes: referring to cloud hosts in the same weak anti-affinity group, preferably distributed on different physical nodes, when available physical nodes are insufficient, policies are no longer followed at this time, and the cloud hosts may be distributed on the same physical node.
Secondly, an application scenario involved in the embodiment of the present invention is described.
OpenStack cloud host scheduling refers to the process of distributing cloud host instances onto appropriate physical hosts in an OpenStack cloud computing platform. The purpose of OpenStack cloud host scheduling is to achieve optimal utilization of resources to improve performance and reliability of the overall cloud computing environment.
The OpenStack default scheduler is a filter and weighting algorithm based scheduler. The method can filter and sort physical hosts according to performance parameter conditions and affinity group strategies appointed by users, select the best physical host to deploy cloud host examples, and achieve scheduling of the cloud hosts.
The prior art has the following defects:
(1) In an actual application scene, in the process of dispatching an affinity group cloud host, binding of the affinity group can only be performed when the cloud host is created, and then the affinity group of the cloud host cannot be bound or unbind, so that the cloud host dispatching process is not flexible enough, and the adaptability of the existing cloud host dispatching method is poor;
(2) When the weak affinity and weak anti-affinity cloud hosts are scheduled, as a plurality of factors influencing the physical nodes of the weighting device are selected, besides being related to the weak affinity and weak anti-affinity, the influencing factors also comprise memory utilization rate, CPU utilization rate, IO load of the whole system and the like, so that the situation that the weak affinity and weak anti-affinity cloud hosts are failed to schedule often occurs.
Based on the defects, the embodiment of the invention provides the cloud platform with stronger adaptability, high server resource utilization rate and strong system reliability, and solves the problem that weak affinity and weak anti-affinity scheduling can fail in the cloud platform.
The solution of cloud host scheduling provided by the present invention is described below with reference to fig. 1-5.
Fig. 1 is a schematic flow chart of a cloud host scheduling method provided by the present invention, and as shown in fig. 1, the method includes:
step 101, receiving a first request for scheduling a cloud host binding an affinity group;
specifically, cloud host affinity refers to a technology of deploying multiple instances of the same application on the same physical host or the same physical server cluster in order to improve the reliability and performance of the application in a cloud computing environment. Alternatively, the binding or unbinding of the affinity group and the cloud host in this embodiment may be performed at any time when the state of the cloud host is the off state.
When the cloud host of the binding affinity group needs to be scheduled, the first request is received, for example, the cloud host of the binding affinity group is scheduled, where the process of scheduling the cloud host may be understood as a process of distributing a cloud host instance to an appropriate physical host (physical node) in the cloud computing platform, where the cloud host instance is for example, downloading a mirror image, configuring a network security group (in and out), creating a (cloud) host, creating a key pair (key pair), and the like.
Step 102, in response to the first request, judging whether the first node meets a preset first strategy; the first strategy is an affinity group strategy determined based on the cloud host;
specifically, after receiving the first request, in response to the first request for scheduling the cloud host binding the affinity group, that is, receiving a request for distributing the cloud host instance to an appropriate physical host (node), it is first determined whether the first node meets a first policy, where the first node is a current node where the cloud host instance is located, and the first policy is an affinity group policy determined based on the cloud host. Among them, affinity group policies include, for example, node affinities, node anti-affinities, node weak affinities, and node weak anti-affinities. Node affinity means that cloud hosts in the same affinity group must coexist on the same node; node affinity refers to the fact that cloud hosts in the same affinity group must be distributed across different nodes; node weak affinity refers to cloud hosts in the same weak affinity group, preferably coexist on the same node, and when the resources of the node are insufficient, no policy is followed at this time, and the cloud hosts can be distributed on other nodes; node weak anti-affinity refers to cloud hosts in the same weak anti-affinity group, preferably distributed on different nodes, when available nodes are insufficient, no policy is followed at this time, and cloud hosts may be distributed on the same node.
Further, determining whether the first node meets the preset first policy may be understood as determining whether the current node where the cloud host instance is located is bound to an affinity group. It is understood that the first node satisfies the first policy (affinity group policy), that is to say that the first node (current physical node) binds the affinity group.
Step 103, under the condition that the first node does not meet the first policy, filtering the plurality of second nodes based on the performance parameter information to obtain at least one alternative node;
specifically, after passing through step 102, the results that can be obtained include two types:
when the first node meets a preset first strategy, namely the first node binds an affinity group, starting the cloud host on the first node; when the first node does not meet the preset first policy, that is, the first node is not bound to the affinity group, the target node to be migrated needs to be further acquired. The procedure for obtaining the target node to be migrated is exemplified as follows:
first, the plurality of second nodes may be filtered by a specified filter based on performance parameter information, such as memory usage, CPU occupancy, disk space, etc., to obtain at least one candidate node. Optionally, multiple filters may be used to perform screening sequentially to obtain at least one candidate node meeting the necessary condition, for example, first, a first filter is used to filter to obtain a second node where the available quota of the memory meets the requirements of the cloud host to be migrated, further, screening is further performed through the second filter in the second node where the available quota of the memory meets the requirements of the cloud host to be migrated, so as to obtain a second node where the available quota of the CPU meets the requirements of the cloud host to be migrated, and by analogy, at least one candidate node meeting the necessary condition may be screened.
104, obtaining a sequencing result according to at least one alternative node, weight information corresponding to each alternative node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host;
specifically, after at least one candidate node satisfying the necessary condition is obtained in step 103, weight information corresponding to each candidate node may be further calculated, and then the candidate nodes are ranked based on a preset second policy, so as to obtain a ranking result, where the second policy includes a weak affinity policy determined based on the cloud host and/or a weak inverse affinity policy determined based on the cloud host, and the weight information corresponding to each candidate node indicates a probability that the weighting device selects the candidate node.
Step 105, determining a target node to be migrated according to the sequencing result;
specifically, after the sorting result of each candidate node is obtained, the target result to be migrated may be determined based on the sorting result. For example, the candidate node corresponding to the maximum weight value in the sorting result is determined as the target node.
And 106, migrating the cloud host to the target node before the cloud host is started.
Specifically, after determining the target node, the cloud host may be further migrated to the target node, for example, an instance of the cloud host is allocated to the target node. It should be specifically noted that, in the embodiment of the present invention, the sequence of starting up and migration of the cloud host is that the cloud host is migrated to the target node first, and then the cloud host is started up.
Optionally, the cloud host and the affinity group may be bound, for example, or the cloud host and the affinity group may be unbound, at any time when the state of the cloud host is in the shutdown state, so that the overall scheduling process has higher flexibility.
In the method provided by the embodiment, by receiving a first request for scheduling the cloud host binding the affinity group, responding to the first request, and judging whether the first node meets a preset first strategy, wherein the first strategy is an affinity group strategy determined based on the cloud host; further, under the condition that the first node does not meet the first policy, filtering the plurality of second nodes based on the performance parameter information to obtain at least one alternative node; then, according to at least one alternative node, weight information corresponding to each alternative node and a preset second strategy, a sequencing result is obtained, wherein the second strategy comprises at least one of a weak affinity strategy determined based on a cloud host and a weak anti-affinity strategy determined based on the cloud host; determining a target node to be migrated according to the sequencing result; and finally, migrating the cloud host to the target node before the cloud host is started. According to the method, when the first node does not meet the preset first strategy, the plurality of second nodes are filtered based on the performance parameter information to obtain at least one alternative node, and then, according to the at least one alternative node, the weight information corresponding to each alternative node and the preset second strategy, a sequencing result is obtained, the target node to be migrated is determined according to the sequencing result, finally, the cloud host is migrated to the target node before the cloud host is started, and the cloud host scheduling process based on the first strategy and the second strategy is put before the cloud host is started, so that flexibility in cloud host scheduling is improved, and the method is more suitable for actual use scenes.
Optionally, according to at least one candidate node, weight information corresponding to each candidate node, and a preset second policy, a ranking result is obtained, including:
judging whether at least one alternative node meets a second strategy;
and under the condition that the at least one alternative node meets the second strategy, the at least one alternative node is weighted and sequenced by utilizing the weight information corresponding to the at least one alternative node, so as to obtain a sequencing result.
Specifically, in step 104, it may first be determined whether at least one candidate node satisfies a second policy, where the second policy includes a weak affinity policy determined based on the cloud host and/or a weak anti-affinity policy determined based on the cloud host.
Further, in the case that at least one of the candidate nodes satisfies the second policy, that is, at least one of the candidate nodes satisfies the weak affinity policy, or satisfies the weak anti-affinity policy, or satisfies the weak affinity policy and the weak anti-affinity policy, the weight of at least one of the candidate nodes that satisfies the second policy may be calculated by using a weight device, where the calculation basis of the weight device includes the system CPU occupancy rate, the system memory usage rate, and the input/output IO load of the whole machine, so as to obtain the weight information corresponding to each of the candidate nodes. The process of obtaining the weight information corresponding to each candidate node is exemplified as follows:
The weight allocation of the weighter is that the weight of the system CPU occupancy rate is 30 percent, the weight of the system memory usage rate is 30 percent, and the weight of the input-output IO load of the whole machine is 20 percent, and three alternative nodes are an alternative node 1, an alternative node 2 and an alternative node 3. The weight information corresponding to each candidate node is as follows:
the weight value of the alternative node 1 is 50;
the weight value of the alternative node 2 is 80;
the weight value of the candidate node 3 is 90.
In the method provided by the embodiment, whether the at least one alternative node meets a second strategy is judged, wherein the second strategy is at least one of a weak affinity strategy determined based on the cloud host and a weak anti-affinity strategy determined based on the cloud host, and under the condition that the at least one alternative node meets the second strategy, the at least one alternative node is weighted and sequenced by utilizing weight information corresponding to the at least one alternative node, so that a sequencing result is obtained; further, the target node to be migrated may be determined based on the ordering result. In this embodiment, the judging module whether to satisfy the weak affinity or weak inverse affinity policy is extracted separately, and is placed after the filter and before the weighting device to be implemented as an independent module, so that the problem of high failure probability of weak affinity scheduling is solved, and optimal utilization of server resources is achieved.
Optionally, before determining whether the first node meets the preset first policy in response to the first request, the method further includes:
adding a distributed lock to an affinity group corresponding to the cloud host, wherein the distributed lock is used for indicating the migration serial execution of the cloud hosts in the same affinity group; the distributed lockset is provided with a unique KEY value KEY; the affinity group corresponding to the cloud host has a unique identification ID.
Specifically, the distributed lock includes three implementations: 1. implementing a distributed lock based on a database; 2. implementing a distributed lock based on a cache (Redis) manner; 3. a distributed lock is implemented based on a Zookeeper component. From a performance perspective (from high to low): the cache mode is superior to the Zookeeper mode, and the Zookeeper mode is superior to the database mode.
Because the migration and the starting process of the cloud host need to consume a certain time, in order to increase the reliability of the whole system, a distributed lock can be added for the affinity group in the migration and the starting processes of the cloud host, wherein the distributed lock has a unique KEY value KEY, and the affinity group corresponding to the cloud host has a unique identification ID. In the embodiment of the invention, a distributed lock is added for the affinity group, namely, the locks added by all cloud hosts in the same affinity group are the same lock, and the advantage of the distributed lock is that the cloud hosts in the same affinity group can only be executed in series in the migration process, that is, one migration process can only execute the migration of one cloud host in the same affinity group, the other cloud hosts in the same affinity group are blocked, and the migration is performed after the lock is obtained by other cloud hosts after the migration of the cloud hosts currently being migrated is finished, and the lock is released. For cloud hosts of different affinity groups, because the IDs of the affinity groups are different, the added locks are not identical and do not affect each other.
In this embodiment, the distributed lock is implemented, for example, by invoking the nova.incorporation component provided by OpenStack, and the underlying cache-based Redis. Fig. 2 is a second flow chart of the cloud host scheduling method provided by the present invention, as shown in fig. 2, an example of a cloud host migration and startup process based on a distributed lock is as follows:
step 201, receiving a first request for migrating a cloud host bound with an affinity group;
step 202, adding a distributed lock to an affinity group corresponding to a cloud host;
step 203, judging whether the first node meets an affinity group policy corresponding to the cloud host;
when the first node meets the preset affinity group policy, executing the following steps:
step 204, starting up the cloud host at the first node;
when the first node does not meet the preset affinity group policy, executing the following steps:
step 205, obtaining a target node to be migrated by using a scheduler;
step 206, judging whether the target node to be migrated meets a weak affinity or weak anti-affinity strategy;
when the target node to be migrated does not meet the weak affinity or weak anti-affinity policy, executing the following steps:
step 204, starting up the cloud host at the first node;
when the target node to be migrated meets the weak affinity or weak anti-affinity strategy, executing the following steps:
Step 207, migrating the cloud host to a target node, and starting up the cloud host at the target node;
step 208, releasing the distributed lock.
Specifically, a first request for migrating the cloud host bound with the affinity group is received first, and a distributed lock is added to the affinity group corresponding to the cloud host.
Further, judging whether the first node meets an affinity group strategy corresponding to the cloud host, and starting up the cloud host at the first node when the first node meets a preset affinity group strategy; when the first node does not meet a preset affinity group strategy, a target node to be migrated is obtained by using a scheduler; judging whether the target node to be migrated meets a weak affinity or weak anti-affinity strategy, and starting the cloud host at the first node when the target node to be migrated does not meet the weak affinity or weak anti-affinity strategy; and when the target node to be migrated meets the weak affinity or weak anti-affinity strategy, migrating the cloud host to the target node, and starting the cloud host at the target node.
And finally, the distributed lock is released after the migration and the startup of the cloud host are completed.
In the method provided by the embodiment, the implementation of the whole method is based on an OpenStack cloud computing platform, the situation that the acquired data are inconsistent and the final dispatching result is wrong due to the fact that a plurality of processes access the same affinity group in a distributed environment during affinity dispatching is avoided through a distributed lock, and the reliability of the whole distributed system is improved through the distributed lock.
Optionally, after the cloud host is migrated to the target node when the cloud host is powered on, the method further includes:
releasing the distributed lock.
Specifically, the time for releasing the distributed lock may be when the cloud host is migrated to the target node, and the distributed lock is released after the target node boots the cloud host. Or when the first node meets the preset affinity group policy, the distributed lock is released after the first node starts up the cloud host. As shown in fig. 2, the operation procedure of "step 208, release the distributed lock" corresponds to this.
In the method provided by the embodiment, the reliability of the whole distributed system is improved by utilizing the distributed lock.
Optionally, determining the target node to be migrated according to the sorting result includes:
determining an alternative node corresponding to the maximum weight value in the sequencing result as a target node; or alternatively, the first and second heat exchangers may be,
according to the number K of the preset target nodes and the sequencing result, K candidate nodes with the maximum weight value are sequentially determined in at least one candidate node to be the target nodes; k is a positive integer.
Specifically, after the sorting result is obtained, that is, after the weight information corresponding to each candidate node is obtained, for example, the weight value corresponding to each candidate node, the manner of determining the target node to be migrated is exemplified as follows:
In a first mode, the weight values corresponding to the candidate nodes are ordered, and the candidate node with the largest weight value in the ordering result is determined to be the target node.
Specifically, it is assumed that there are multiple candidate nodes in the obtained sorting result, such as candidate node 1, candidate node 2, and candidate node 3, and each candidate node has a corresponding weight value, such as a weight value of 50 for candidate node 1, a weight value of 80 for candidate node 2, and a weight value of 90 for candidate node 3. In the first embodiment, the candidate node corresponding to the maximum weight value, that is, the candidate node 3, may be determined as the target node.
Mode two: and sequentially determining K candidate nodes with the maximum weight value as target nodes in a plurality of candidate nodes according to the number K of the preset target nodes and the sequencing result, wherein K is a positive integer.
Specifically, it is assumed that there are multiple candidate nodes in the obtained sorting result, such as candidate node 1, candidate node 2, and candidate node 3, and each candidate node has a corresponding weight value, such as a weight value of 50 for candidate node 1, a weight value of 80 for candidate node 2, and a weight value of 90 for candidate node 3. In the second mode, the number K of target nodes may be preset in advance, for example, k=2, and the candidate nodes corresponding to the first 2 weight values in the weight value sequence may be determined as target nodes, that is, the candidate node 2 and the candidate node 3.
In the method provided by the embodiment of the invention, the weight values of a plurality of candidate nodes are sequenced, and then the candidate node with the largest weight value is determined as the target node; or sequentially determining K candidate nodes with the maximum weight value from the plurality of candidate nodes as target nodes according to the preset number K of the target nodes and the sequencing result of the weight value. Because the weight information corresponding to each candidate node represents the probability that the weight device elects the physical node, the target node to be migrated is determined after the weight values are ordered, and the utilization rate of the cloud host to the server resource is higher.
Alternatively, the migration is cold migration.
Specifically, the cloud host is migrated to the target node in step 106, where the migration refers to cold migration. It can be understood that cold migration (cold migration), also called static migration, refers to migration of a virtual machine instance, i.e., a cloud host instance, that is, a cloud host that is in a down state, i.e., shuts down a power supply, when migrating to another physical node. Through cold migration, an associated disk can be selected to be moved from one data storage to another, an instance needs to be restarted to work, the instance is powered down during cold migration, an image file of instance is copied, libvirt.xml is rebuilt (the configuration file relates to network environment, storage device and the like), and the instance is restarted on a target node.
In the method provided by the embodiment, the cloud host is migrated to the target node through cold migration, and the cloud host does not need to be located on the shared memory, so that the data loss rate is small.
Optionally, the method further comprises:
and performing target operation on the cloud host and the affinity group at any moment when the state of the cloud host is in a shutdown state, wherein the target operation comprises at least one of binding the cloud host and the affinity group and unbinding the cloud host and the affinity group.
Specifically, in this embodiment, the scheduling of the cloud host binding the affinity group is completed before the cloud host starts up, and the target operation can be performed on the cloud host and the affinity group at any time when the state of the cloud host is in the shutdown state, where the target operation includes binding and/or unbinding the cloud host and the affinity group.
In the method provided by the embodiment, the affinity group and the cloud hosts are bound and unbindd at any time, so that the cloud hosts which are already created before can be reused without recreating the cloud hosts for realizing the affinity between the cloud hosts. The relationship between the affinity group and the cloud host becomes flexible and is more practical in a real scene.
Fig. 3 is a third flow chart of the cloud host scheduling method provided by the present invention, and as shown in fig. 3, a process for obtaining a target node to be migrated by using a scheduler is illustrated as follows:
step 301, filtering the plurality of second nodes based on the performance parameter information by using a filter to obtain alternative nodes;
step 302, judging whether the candidate node is a weak affinity or a weak anti-affinity group;
when the candidate node is not a weak affinity group or a weak anti-affinity group, determining that the candidate node does not satisfy the weak affinity or weak anti-affinity policy;
when the candidate node is a weak affinity or weak anti-affinity group, the following steps are performed:
step 303, screening to obtain at least one alternative node meeting weak affinity or weak anti-affinity strategies;
step 304, carrying out weight sorting on at least one alternative node by using a weight device to obtain a sorting result;
step 305, determining the target node to be migrated according to the sorting result.
Specifically, the process of obtaining the target node to be migrated by using the scheduler is as follows: first, a plurality of second nodes are filtered based on performance parameter information by using a filter, and an alternative node is obtained.
Further, judging whether the candidate node is a weak affinity group or a weak anti-affinity group, and when the candidate node is not the weak affinity group or the weak anti-affinity group, determining that the candidate node does not meet the weak affinity or the weak anti-affinity strategy; when the candidate nodes are weak affinity or weak anti-affinity groups, screening to obtain at least one candidate node meeting the weak affinity or weak anti-affinity strategy; and further, weighting and sorting at least one alternative node by using a weighting device to obtain a sorting result.
And finally, determining the target node to be migrated according to the sequencing result. For example, the candidate node corresponding to the maximum weight value in the sorting result is determined as the target node, and scheduling is completed based on the target node.
The cloud host scheduling device provided by the invention is described below, and the cloud host scheduling device and the cloud host scheduling method described below can be referred to correspondingly.
Fig. 4 is a schematic structural diagram of a cloud host scheduling device provided by the present invention, as shown in fig. 4, the device includes:
a receiving module 410, configured to receive a first request for scheduling a cloud host binding an affinity group;
the scheduling module 420 is configured to determine, in response to the first request, whether the first node meets a preset first policy; the first policy is an affinity group policy determined based on the cloud host; filtering a plurality of second nodes based on performance parameter information under the condition that the first node does not meet the first policy to obtain at least one alternative node; obtaining a sequencing result according to the weight information corresponding to each candidate node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host; determining a target node to be migrated according to the sequencing result;
And the migration module 430 migrates the cloud host to the target node before the cloud host is started.
In the device provided in this embodiment, the receiving module 410 receives a first request for scheduling the cloud host binding the affinity group, and the scheduling module 420 determines, in response to the first request, whether the first node meets a preset first policy, where the first policy is an affinity group policy determined based on the cloud host; further, under the condition that the first node does not meet the first policy, filtering the plurality of second nodes based on the performance parameter information to obtain at least one alternative node; then, according to at least one alternative node, weight information corresponding to each alternative node and a preset second strategy, a sequencing result is obtained, wherein the second strategy comprises at least one of a weak affinity strategy determined based on a cloud host and a weak anti-affinity strategy determined based on the cloud host; determining a target node to be migrated according to the sequencing result; finally, the migration module 430 migrates the cloud host to the target node before the cloud host is powered on. According to the method, when the first node does not meet the preset first strategy, the plurality of second nodes are filtered based on the performance parameter information to obtain at least one alternative node, and then, according to the at least one alternative node, the weight information corresponding to each alternative node and the preset second strategy, a sequencing result is obtained, the target node to be migrated is determined according to the sequencing result, finally, the cloud host is migrated to the target node before the cloud host is started, and the cloud host scheduling process based on the first strategy and the second strategy is put before the cloud host is started, so that flexibility in cloud host scheduling is improved, and the method is more suitable for actual use scenes.
Optionally, the scheduling module 420 is specifically configured to:
judging whether the at least one alternative node meets the second strategy;
and under the condition that at least one candidate node meets the second strategy, carrying out weight ranking on the at least one candidate node by utilizing weight information corresponding to each at least one candidate node, and obtaining the ranking result.
Optionally, the cloud host scheduling apparatus 400 further includes a distributed locking module for:
adding a distributed lock to the affinity group corresponding to the cloud host, wherein the distributed lock is used for indicating the migration serial execution of the cloud hosts in the same affinity group; the distributed lockset is provided with a unique KEY value KEY; and the affinity group corresponding to the cloud host has a unique identification ID.
Optionally, the cloud host scheduling apparatus 400 further includes a distributed lock unlocking module, configured to:
releasing the distributed lock.
Optionally, the scheduling module 420 is further configured to:
determining an alternative node corresponding to the maximum weight value in the sequencing result as the target node; or alternatively, the first and second heat exchangers may be,
according to the number K of the preset target nodes and the sequencing result, K candidate nodes with the maximum weight value are sequentially determined as the target nodes in the at least one candidate node; k is a positive integer.
Optionally, the migration is cold migration.
Optionally, the cloud host scheduling apparatus 400 further includes an affinity group binding module, configured to:
and performing target operation on the cloud host and the affinity group at any moment when the state of the cloud host is in a shutdown state, wherein the target operation comprises at least one of binding the cloud host and the affinity group and unbinding the cloud host and the affinity group.
Fig. 5 illustrates a physical schematic diagram of an electronic device, as shown in fig. 5, which may include: processor 510, communication interface (Communications Interface) 520, memory 530, and communication bus 540, wherein processor 510, communication interface 520, memory 530 complete communication with each other through communication bus 540. Processor 510 may invoke logic instructions in memory 530 to perform a cloud host scheduling method comprising:
receiving a first request for scheduling the cloud hosts binding the affinity group;
responding to the first request, and judging whether the first node meets a preset first strategy; the first policy is an affinity group policy determined based on the cloud host;
Filtering a plurality of second nodes based on performance parameter information under the condition that the first node does not meet the first policy to obtain at least one alternative node;
obtaining a sequencing result according to the weight information corresponding to each candidate node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host;
determining a target node to be migrated according to the sequencing result;
and before the cloud host is started, migrating the cloud host to the target node.
Further, the logic instructions in the memory 530 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a 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, a network device, etc.) to perform all or part of the steps of the method of the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product, where the computer program product includes a computer program, where the computer program can be stored on a non-transitory computer readable storage medium, and when the computer program is executed by a processor, the computer can execute the cloud host scheduling method provided by the above methods, and the method includes:
receiving a first request for scheduling the cloud hosts binding the affinity group;
responding to the first request, and judging whether the first node meets a preset first strategy; the first policy is an affinity group policy determined based on the cloud host;
filtering a plurality of second nodes based on performance parameter information under the condition that the first node does not meet the first policy to obtain at least one alternative node;
obtaining a sequencing result according to the weight information corresponding to each candidate node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host;
determining a target node to be migrated according to the sequencing result;
And before the cloud host is started, migrating the cloud host to the target node.
In yet another aspect, the present invention further provides a non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor is implemented to perform the cloud host scheduling method provided by the above methods, the method comprising:
receiving a first request for scheduling the cloud hosts binding the affinity group;
responding to the first request, and judging whether the first node meets a preset first strategy; the first policy is an affinity group policy determined based on the cloud host;
filtering a plurality of second nodes based on performance parameter information under the condition that the first node does not meet the first policy to obtain at least one alternative node;
obtaining a sequencing result according to the weight information corresponding to each candidate node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host;
determining a target node to be migrated according to the sequencing result;
And before the cloud host is started, migrating the cloud host to the target node.
The apparatus embodiments described above are merely illustrative, wherein elements illustrated as separate elements may or may not be physically separate, and elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solutions may be embodied essentially or in part in the form of a software product, which may be stored in a computer-readable storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the various embodiments or methods of some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. The cloud host scheduling method is characterized by comprising the following steps of:
receiving a first request for scheduling the cloud hosts binding the affinity group;
responding to the first request, and judging whether the first node meets a preset first strategy; the first policy is an affinity group policy determined based on the cloud host;
filtering a plurality of second nodes based on performance parameter information under the condition that the first node does not meet the first policy to obtain at least one alternative node;
obtaining a sequencing result according to the weight information corresponding to each candidate node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host;
Determining a target node to be migrated according to the sequencing result;
and before the cloud host is started, migrating the cloud host to the target node.
2. The cloud host scheduling method according to claim 1, wherein the obtaining a ranking result according to the at least one candidate node, the weight information corresponding to each candidate node, and a preset second policy includes:
judging whether the at least one alternative node meets the second strategy;
and under the condition that at least one candidate node meets the second strategy, carrying out weight ranking on the at least one candidate node by utilizing weight information corresponding to each at least one candidate node, and obtaining the ranking result.
3. The cloud host scheduling method according to claim 1 or 2, wherein before the determining, in response to the first request, whether the first node meets a preset first policy, the method further includes:
adding a distributed lock to the affinity group corresponding to the cloud host, wherein the distributed lock is used for indicating the migration serial execution of the cloud hosts in the same affinity group; the distributed lockset is provided with a unique KEY value KEY; and the affinity group corresponding to the cloud host has a unique identification ID.
4. The cloud host scheduling method of claim 3, further comprising, after migrating the cloud host to a target node when the cloud host is powered on:
releasing the distributed lock.
5. The cloud host scheduling method according to claim 1 or 2, wherein the determining the target node to be migrated according to the sorting result includes:
determining an alternative node corresponding to the maximum weight value in the sequencing result as the target node; or alternatively, the first and second heat exchangers may be,
according to the number K of the preset target nodes and the sequencing result, K candidate nodes with the maximum weight value are sequentially determined as the target nodes in the at least one candidate node; k is a positive integer.
6. The cloud host scheduling method of claim 1 or 2, wherein the migration is a cold migration.
7. The cloud host scheduling method according to claim 1 or 2, characterized in that the method further comprises:
and performing target operation on the cloud host and the affinity group at any moment when the state of the cloud host is in a shutdown state, wherein the target operation comprises at least one of binding the cloud host and the affinity group and unbinding the cloud host and the affinity group.
8. A cloud host scheduling apparatus, comprising:
the receiving module is used for receiving a first request for scheduling the cloud host binding the affinity group;
the scheduling module is used for responding to the first request and judging whether the first node meets a preset first strategy or not; the first policy is an affinity group policy determined based on the cloud host; filtering a plurality of second nodes based on performance parameter information under the condition that the first node does not meet the first policy to obtain at least one alternative node; obtaining a sequencing result according to the weight information corresponding to each candidate node and a preset second strategy; the second policy includes at least one of a weak affinity policy determined based on the cloud host and a weak anti-affinity policy determined based on the cloud host; determining a target node to be migrated according to the sequencing result;
and the migration module migrates the cloud host to the target node before the cloud host is started.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the cloud host scheduling method of any of claims 1 to 7 when the program is executed by the processor.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the cloud host scheduling method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311190639.6A CN116938943B (en) | 2023-09-15 | 2023-09-15 | Cloud host scheduling method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311190639.6A CN116938943B (en) | 2023-09-15 | 2023-09-15 | Cloud host scheduling method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116938943A true CN116938943A (en) | 2023-10-24 |
| CN116938943B CN116938943B (en) | 2024-01-12 |
Family
ID=88382945
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311190639.6A Active CN116938943B (en) | 2023-09-15 | 2023-09-15 | Cloud host scheduling method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116938943B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3244309A1 (en) * | 2016-05-12 | 2017-11-15 | Alcatel Lucent | Method, system and computer readable medium to update scheduling hints |
| CN110389814A (en) * | 2019-06-28 | 2019-10-29 | 苏州浪潮智能科技有限公司 | A kind of cloud host migration dispatching method, system, terminal and storage medium |
| US20200004582A1 (en) * | 2018-07-02 | 2020-01-02 | International Business Machines Corporation | Cognitive cloud migration optimizer |
| CN110727512A (en) * | 2019-09-30 | 2020-01-24 | 星环信息科技（上海）有限公司 | Cluster resource scheduling method, device, equipment and storage medium |
| CN111190688A (en) * | 2019-12-19 | 2020-05-22 | 西安交通大学 | Cloud data center-oriented Docker migration method and system |
| CN112394881A (en) * | 2020-11-04 | 2021-02-23 | 华云数据控股集团有限公司 | Cloud host migration method and electronic equipment |
| CN114020405A (en) * | 2021-10-27 | 2022-02-08 | 郑州云海信息技术有限公司 | Virtual machine scheduling method, device, equipment and medium |
| WO2022104396A1 (en) * | 2020-11-16 | 2022-05-19 | Juniper Networks, Inc. | Active assurance for virtualized services |
| CN115658301A (en) * | 2022-10-20 | 2023-01-31 | 中国电信股份有限公司 | Storage resource scheduling method, device, storage medium and electronic equipment |
| CN116208613A (en) * | 2023-02-13 | 2023-06-02 | 天翼云科技有限公司 | Migration method and device of cloud host, electronic equipment and storage medium |
| CN116233144A (en) * | 2023-02-03 | 2023-06-06 | 上海德拓信息技术股份有限公司 | Method and system for forcing batch evacuation of same-node cloud hosts in cloud scene |
-
2023
- 2023-09-15 CN CN202311190639.6A patent/CN116938943B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3244309A1 (en) * | 2016-05-12 | 2017-11-15 | Alcatel Lucent | Method, system and computer readable medium to update scheduling hints |
| US20200004582A1 (en) * | 2018-07-02 | 2020-01-02 | International Business Machines Corporation | Cognitive cloud migration optimizer |
| CN110389814A (en) * | 2019-06-28 | 2019-10-29 | 苏州浪潮智能科技有限公司 | A kind of cloud host migration dispatching method, system, terminal and storage medium |
| CN110727512A (en) * | 2019-09-30 | 2020-01-24 | 星环信息科技（上海）有限公司 | Cluster resource scheduling method, device, equipment and storage medium |
| CN111190688A (en) * | 2019-12-19 | 2020-05-22 | 西安交通大学 | Cloud data center-oriented Docker migration method and system |
| CN112394881A (en) * | 2020-11-04 | 2021-02-23 | 华云数据控股集团有限公司 | Cloud host migration method and electronic equipment |
| WO2022104396A1 (en) * | 2020-11-16 | 2022-05-19 | Juniper Networks, Inc. | Active assurance for virtualized services |
| CN114020405A (en) * | 2021-10-27 | 2022-02-08 | 郑州云海信息技术有限公司 | Virtual machine scheduling method, device, equipment and medium |
| CN115658301A (en) * | 2022-10-20 | 2023-01-31 | 中国电信股份有限公司 | Storage resource scheduling method, device, storage medium and electronic equipment |
| CN116233144A (en) * | 2023-02-03 | 2023-06-06 | 上海德拓信息技术股份有限公司 | Method and system for forcing batch evacuation of same-node cloud hosts in cloud scene |
| CN116208613A (en) * | 2023-02-13 | 2023-06-02 | 天翼云科技有限公司 | Migration method and device of cloud host, electronic equipment and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| DINGKUN YANG等: "Provisioning with Fine-grained Affinity for Container-enabled Cloud-edge System", 《2021 IEEE INTL CONF ON PARALLEL & DISTRIBUTED PROCESSING WITH APPLICATIONS, BIG DATA & CLOUD COMPUTING, SUSTAINABLE COMPUTING & COMMUNICATIONS, SOCIAL COMPUTING & NETWORKING (ISPA/BDCLOUD/SOCIALCOM/SUSTAINCOM)》, pages 1675 - 1681 * |
| EDZ 2000: "openstack 冷迁移实例", Retrieved from the Internet <URL:https://blog.csdn.net/EDISON_ZHOU/article/details/107516689#> * |
| LUKA_JA: "OpenStack nova-scheduler调度过程", Retrieved from the Internet <URL:https://blog.csdn.net/luuJa_IQ/article/details/104118146> * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116938943B (en) | 2024-01-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109947557B (en) | Distributed lifecycle management for cloud platforms | |
| US11960915B2 (en) | Method and apparatus for creating virtual machine based on parameter information of a virtual network interface card | |
| CN110647394B (en) | Resource allocation method, device and equipment | |
| CN113918270A (en) | Cloud resource scheduling method and system based on Kubernetes | |
| CN105700939A (en) | Method and system for multi-thread synchronization in distributed system | |
| Mohamed et al. | Experiences with the koala co-allocating scheduler in multiclusters | |
| CN110838939B (en) | Scheduling method based on lightweight container and edge Internet of things management platform | |
| CN111988160B (en) | Method and device for deploying virtualized network function | |
| CN106790092B (en) | Remote procedure call server control system and method | |
| CN112835714B (en) | Container arrangement method, system and medium for CPU heterogeneous clusters in cloud edge environment | |
| CN106445631B (en) | Method and system for deploying virtual machine and physical server | |
| CN114153580A (en) | Cross-multi-cluster work scheduling method and device | |
| CN111857951A (en) | Containerized deployment platform and deployment method | |
| CN112988344A (en) | Distributed batch task scheduling method, device, equipment and storage medium | |
| CN106789308A (en) | The GIS service device and its control method of a kind of micro services framework automatically retractable | |
| CN114253459A (en) | Method and device for creating persistent data volume and server | |
| JP6010975B2 (en) | Job management apparatus, job management method, and program | |
| CN110034963B (en) | Application cluster self-adaptive elastic configuration method | |
| CN115964176B (en) | Cloud computing cluster scheduling method, electronic equipment and storage medium | |
| CN116938943B (en) | Cloud host scheduling method, device, equipment and storage medium | |
| CN105827744A (en) | Data processing method of cloud storage platform | |
| CN111240824A (en) | CPU resource scheduling method and electronic equipment | |
| Wu et al. | Abp scheduler: Speeding up service spread in docker swarm | |
| CN116680078A (en) | Cloud computing resource scheduling method, device, equipment and computer storage medium | |
| CN116755849A (en) | Execution method of privacy calculation task, privacy task controller and central dispatcher |
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 |