CN106302632A - The method for down loading of a kind of foundation image and management node - Google Patents
The method for down loading of a kind of foundation image and management node Download PDFInfo
- Publication number
- CN106302632A CN106302632A CN201610586890.8A CN201610586890A CN106302632A CN 106302632 A CN106302632 A CN 106302632A CN 201610586890 A CN201610586890 A CN 201610586890A CN 106302632 A CN106302632 A CN 106302632A
- Authority
- CN
- China
- Prior art keywords
- node
- container
- computing nodes
- foundation image
- target computing
- 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
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
- H04L67/1095—Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
- G06F8/61—Installation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
- G06F8/61—Installation
- G06F8/63—Image based installation; Cloning; Build to order
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Information Transfer Between Computers (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses the method for down loading of a kind of foundation image, described method is implemented in the management node in container cluster management framework, described container cluster management framework also includes far-end mirror image warehouse and at least one calculating node, described method includes: obtain the status information of each described calculating node and the use information of container that each described calculating node once started, described use information include described in the used foundation image of container that once started;From all of described calculating node, filter out its status information meet pre-conditioned target computing nodes;Sending pre-download order to described target computing nodes, described pre-download order downloads the mark of the used foundation image of container that described target computing nodes once started in advance for indicating described target computing nodes from described far-end mirror image warehouse.Accordingly, the embodiment of the invention also discloses a kind of management node.Use the present invention, it is possible to achieve accelerate the startup time of container.
Description
Technical field
The present invention relates to field of computer technology, particularly relate to method for down loading and the management node of a kind of foundation image.
Background technology
Intel Virtualization Technology is a kind of resource management techniques being applied to computer, can be by the various physical resources of computer
(such as network, internal memory and storage) give abstract after present, broken the boundary between physical arrangement, it is achieved that physics provide
Maximally utilizing of source.Along with the progressively maturation of traditional virtual technology, performance and resource utilization are already close to bottleneck, therefore
Propose in the industry a kind of New Virtual technology being referred to as container technique.Comparing traditional virtual technology, container technique has money
Source loss is less, resource utilization is higher and the faster advantage of the speed of service.
Referring to Fig. 9, main frame, when starting a container, need to first determine the foundation image that container to be launched is corresponding, connect
And download this foundation image from far-end mirror image warehouse, then use this foundation image to start container.But, when foundation image
Size is relatively big, or when speed of download is slower, the time starting container will be greatly prolonged.
Summary of the invention
Embodiments provide method for down loading and the management node of a kind of foundation image, it is possible to achieve accelerate container
The startup time.
Embodiment of the present invention first aspect provides the method for down loading of a kind of foundation image, and the method is implemented in container cluster
Management node in management framework, container cluster management framework also includes far-end mirror image warehouse and at least one calculating node, should
Method includes: obtains the use information that each status information calculating node calculates, with each, the container that node once started, makes
The used foundation image of container once started is included by information;Its status information is filtered out full from all of calculating node
The target computing nodes that foot is pre-conditioned;Sending pre-download order to target computing nodes, pre-download order includes once starting
The used foundation image of container, pre-download order is used for indicating target computing nodes to download once from far-end mirror image warehouse in advance
The used foundation image of container started.Calculating node can be allowed to download foundation image in advance, in order to follow-up direct use
The foundation image downloaded in advance starts container, accelerates the startup time of container.
In the first possible implementation of first aspect, use information also includes that the container once started is used
The use frequency of foundation image.Accordingly, control node from all of calculating node, filter out the satisfied default bar of operation conditions
Before the target computing nodes of part, also perform: according to the uses frequency order from high to low of foundation image, each is calculated and saves
The used foundation image of container that point once started is ranked up;Pre-download order also includes that the container once started used
The ranking results of foundation image, pre-download order be used for indicating target computing nodes in advance from far-end mirror image warehouse according to sequence
The used foundation image of container that the progressive download of result once started.Advantage is, calculating node can be allowed preferentially to download
Use the foundation image that frequency is higher, improve the probability that the foundation image downloaded is used.
First aspect the second may in implementation, status information include calculating node processor utilization rate,
Memory usage and local mirror image warehouse utilization rate.Accordingly, control node and filter out its shape from all of calculating node
State information meets the concrete operations of pre-conditioned target computing nodes: from all of calculating node, filters out it and processes
Device utilization rate is less than the first predetermined threshold value, and its memory usage is less than the second predetermined threshold value, and its local mirror image warehouse uses
Rate is less than the target computing nodes of the 3rd predetermined threshold value.Advantage is, status information can be avoided to be unsatisfactory for the calculating joint of condition
Point goes to download in advance foundation image, and then the problem avoiding failed download or over loading.
In conjunction with any one possible implementation in the first to the second of first aspect or first aspect, can at the third
In energy implementation, control node, before target computing nodes sends pre-download order, also performs: determine target computing nodes
Local mirror image warehouse in there is not the used foundation image of container that target computing nodes once started.Advantage is, can
To avoid calculating the foundation image that node download repeats.
In conjunction with any one possible implementation in the first to the second of first aspect or first aspect, can at the 4th kind
In energy implementation, control node, before target computing nodes sends pre-download order, also performs: determine target computing nodes
The most downloading foundation image.Advantage is, can avoid calculating node and download the foundation image repeated.
Embodiment of the present invention second aspect provides a kind of startup method of container, and the method is implemented in container cluster management
Management node in framework, container cluster management framework also includes far-end mirror image warehouse and at least one calculating node, the method
Including: know the specification information of container to be launched and container;Search its status information and meet the operation bar of specification information requirement
The calculating node of part;The target computing nodes being loaded with down foundation image is filtered out from the calculating node searched;To mesh
Mark calculates node and sends container startup order, and container starts the base that order is downloaded the most in advance for indicating target computing nodes use
Plinth image starting container.Can allow the calculating node being loaded with down foundation image to start container, accelerate container startup time
Between.
Second aspect the first may in implementation, processor utilization rate that status information includes calculating node and
Memory usage.Accordingly, the calculating node that its status information of node checks meets the service condition of specification information requirement is managed
Concrete operations be: search its processor utilization rate higher than the first predetermined threshold value, and its memory usage preset higher than second
The calculating node of threshold value.Advantage is, can avoid being unsatisfactory for the calculating node deactivation container of service condition, strengthen feasible
Property.
In the possible implementation of the second of second aspect, control node and filter out the most pre-from the calculating node searched
Before being loaded with down the target computing nodes of foundation image, also perform: the status information calculating node according to each, determine each
Calculate the surplus resources of node.Accordingly, control node filters out from the calculating node searched and is loaded with down basis mirror
The concrete operations of the target computing nodes of picture are: from calculating node of searching filter out its surplus resources the highest the most in advance under
It is loaded with the target computing nodes of foundation image.Advantage is, the calculating node that can preferentially allow surplus resources the highest starts container,
Improve the utilization ratio calculating node.
In conjunction with second aspect the second may implementation, the third may in implementation, control node according to
Each calculates the status information of node, after determining the surplus resources of each calculating node, also performs: if the calculating node searched
In there is no to be loaded with down the target computing nodes of foundation image, the most therefrom filter out the calculating node that its surplus resources is the highest
As target computing nodes.
The embodiment of the present invention third aspect provides the delet method of a kind of foundation image, and the method is implemented in container cluster
Calculating node in management framework, container cluster management framework also includes far-end mirror image warehouse and management node, and the method includes:
The foundation image having not been used in the local mirror image warehouse of inquiry;Status information according to self judges self whether to be in high bearing
Lotus state;The most then delete the foundation image having not been used in local mirror image warehouse.Can allow and calculate knot removal basis mirror
Picture, to reduce load.
In the first possible implementation of the third aspect, status information includes processor utilization rate, memory usage
And local mirror image warehouse utilization rate.Accordingly, calculate node judges self whether to be in high bearing according to the status information of self
The concrete operations of lotus state are: whether decision processor utilization rate is higher than the first predetermined threshold value, and whether memory usage is higher than the
Two predetermined threshold value, and whether local mirror image warehouse utilization rate is higher than the 3rd predetermined threshold value.Advantage is, can allow and be in high load capacity
The calculating knot removal foundation image of state.
In the first possible implementation of the third aspect, calculate in mirror image warehouse, knot removal this locality and have not been used
Foundation image before, also perform: the idle duration of foundation image that inquiry has not been used;During according to foundation image idle
Long order from long to short, is ranked up foundation image.Accordingly, calculating mirror image warehouse, knot removal this locality is not used by
The concrete operations of the foundation image crossed are: delete successively in local mirror image warehouse according to the order of ranking results and have not been used
Foundation image, until self being not at higher load condition.Advantage is, calculating node can be allowed preferentially to delete idle duration
Longer foundation image.
Embodiment of the present invention fourth aspect provides a kind of management node, and described management node has and realizes above-mentioned first party
The behavioral function of the method that face provides, described function can be realized by hardware, it is also possible to performs corresponding software by hardware
Realize.Described hardware or software include one or more module corresponding with above-mentioned functions.
The embodiment of the present invention the 5th aspect provides a kind of management node, and described management node has and realizes above-mentioned second party
The behavioral function of the method that face provides, described function can be realized by hardware, it is also possible to performs corresponding software by hardware
Realize.Described hardware or software include one or more module corresponding with above-mentioned functions.
The embodiment of the present invention the 6th aspect provides and a kind of calculates node, and described management node has and realizes above-mentioned third party
The behavioral function of the method that face provides, described function can be realized by hardware, it is also possible to performs corresponding software by hardware
Realize.Described hardware or software include one or more module corresponding with above-mentioned functions.
The embodiment of the present invention the 7th aspect provides a kind of management node, including communication interface, processor and memorizer, its
In, memorizer stores batch processing, and processor is for calling in memorizer the program of storage, be used for performing following operation:
Obtain the use information of the container that each status information calculating node once started with each calculating node, described
Use information include described in the used foundation image of container that once started;
From all of described calculating node, filter out its status information meet pre-conditioned target computing nodes;
Send pre-download order to described target computing nodes, described pre-download order include described in the container that once started
The mark of used foundation image, described pre-download order is used for indicating described target computing nodes in advance from far-end mirror image storehouse
Once the used foundation image of container started described in the download of storehouse.
Or, it is used for performing following operation:
Know container to be launched and the specification information of described container;
Search the calculating node that its status information meets the service condition of described specification information requirement;
The target computing nodes being loaded with down foundation image is filtered out from the described calculating node searched;
Sending container to described target computing nodes and start order, described container starts order by indicating based on described target
The foundation image downloaded the most in advance described in operator node use starts described container.
Embodiment of the present invention eighth aspect provides a kind of calculating node, including communication interface, processor and memorizer, its
In, memorizer stores batch processing, and processor is for calling in memorizer the program of storage, be used for performing following operation:
The foundation image having not been used in the local mirror image warehouse of inquiry;
Status information according to self judges whether self is in higher load condition;
The most then delete the foundation image having not been used in described local mirror image warehouse.
Therefore, the container cluster management framework in the embodiment of the present invention include manage node, far-end mirror image warehouse with
And at least one calculates node, management node first obtains each status information calculating node and each calculates node and once started
The use information of container, use information includes the used foundation image of container once started, then from all of calculating
Node filters out its status information and meets pre-conditioned target computing nodes, then send pre-download to target computing nodes
Order, pre-download order is used for indicating target computing nodes to download foundation image in advance from far-end mirror image warehouse, in order to target meter
The foundation image that the follow-up direct use of operator node is downloaded in advance starts container, thus realizes accelerating the startup time of container.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention, will use required in embodiment or description of the prior art below
Accompanying drawing be briefly described, it should be apparent that, below describe in accompanying drawing be only some embodiments of the present invention, for this
From the point of view of the those of ordinary skill of field, on the premise of not paying creative work, it is also possible to obtain other according to these accompanying drawings
Accompanying drawing.
Fig. 1 is the schematic flow sheet of the method for down loading of a kind of foundation image that the embodiment of the present invention provides;
Fig. 2 is the schematic flow sheet of the startup method of a kind of container that the embodiment of the present invention provides;
Fig. 3 is the schematic flow sheet of the delet method of a kind of foundation image that the embodiment of the present invention provides;
Fig. 4 is a kind of structural representation managing node that the embodiment of the present invention provides;
Fig. 5 is the structural representation of the another kind of management node that the embodiment of the present invention provides;
Fig. 6 is a kind of structural representation calculating node that the embodiment of the present invention provides;
Fig. 7 is the structural representation of another management node that the embodiment of the present invention provides;
Fig. 8 is the another kind of structural representation calculating node that the embodiment of the present invention provides;
Fig. 9 is the schematic diagram that a kind of foundation image that the embodiment of the present invention provides is downloaded;
Figure 10 is the structural representation of a kind of container cluster management framework that the embodiment of the present invention provides.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Describe, it is clear that described embodiment is only a part of embodiment of the present invention rather than whole embodiments wholely.Based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under not making creative work premise
Embodiment, broadly falls into the scope of protection of the invention.
Figure 10 is the structural representation of a kind of container cluster management framework that the embodiment of the present invention provides, container as shown in the figure
Cluster management framework at least includes managing node, calculating node and far-end mirror image warehouse, and it has been set up communication each other and has connected.
Wherein, management node is the maincenter of whole cluster, main is responsible for the management relevant to container, transmits and dispatch;Calculating node sets
Having the local mirror image warehouse that can store foundation image, main being responsible for uses foundation image to start container;Far-end mirror image warehouse, main
It is responsible for calculating the node Downloadable foundation image of offer.
Fig. 1 is the schematic flow sheet of the method for down loading of a kind of foundation image in the embodiment of the present invention, and the method is implemented in appearance
Management node in device cluster management framework.The flow process of the method for down loading of the foundation image in the present embodiment can be wrapped as shown in the figure
Include:
S101, obtains each status information calculating node and each use calculating the container that node once started is believed
Breath, described use information include described in the used foundation image of container that once started and the use frequency of described foundation image
Rate.
Wherein, status information calculates the running status of node for characterizing, and use information is used for characterizing container to be made in history
Record during with.In the embodiment of the present invention, use information includes calculating the used basis of container that node once started
Mirror image and the use frequency of these foundation image.
It should be understood that, issue an order owing to all of calculating node startup container is required to manage node, therefore manage
Node learns the use information of the container that all of calculating node once started, and these information are saved in the container of management node and adjust
In degree module.
Concrete, management node is connected by calculating the communication of node with each, obtains each state calculating node letter
Breath, and by the record in the local container scheduler module of inquiry, obtains each and calculates making of container that node once started
Use information.
S102, according to the use frequency order from high to low of described foundation image, once opened calculating node each described
The used foundation image of container moved is ranked up.
Such as: assume that the container that calculating node A once started used foundation image x and foundation image y, calculate node B
Once the container started used foundation image x, the container that calculating node C once started to use foundation image z, and assumed again base
The use frequency of plinth mirror image x is 6 times/hour, and the use frequency of foundation image y is 2 times/hour, the use frequency of foundation image z
It is 4 times/hour, then mirror image x → foundation image z → foundation image y based on the ranking results of foundation image.
It is pointed out that the advantage that foundation image is ranked up by management node is, calculating node can be allowed preferential
Download the foundation image using frequency higher, improve the probability that the foundation image downloaded is used.
S103, filters out its status information from all of described calculating node and meets pre-conditioned target calculating joint
Point.
Optionally, described status information includes calculating the processor utilization rate of node, memory usage and local mirror image
Warehouse utilization rate.Wherein, processor utilization rate refers to that CPU (Central Processing Unit, central processing unit) computing provides
The use percentage ratio in source, memory usage refers to that the use percentage ratio of memory source, local mirror image warehouse utilization rate refer to this locality
The storage percentage ratio in mirror image warehouse.It should be understood that, calculate the processor utilization rate of node, memory usage and local mirror image storehouse
Storehouse utilization rate is the highest, illustrates that the load of this calculating node is the highest.
Concrete, management node, from all of calculating node, filters out its processor utilization rate and presets threshold less than first
Value, its memory usage is less than the second predetermined threshold value, and its local mirror image warehouse utilization rate is less than the mesh of the 3rd predetermined threshold value
Mark calculates node.In the embodiment of the present invention, the first predetermined threshold value, the second predetermined threshold value and the 3rd predetermined threshold value can be by managing
Personnel preset, and its value is not especially limited here.Such as: can set the first predetermined threshold value, the second predetermined threshold value and
3rd predetermined threshold value is respectively 60%, 60% and 80%.
It is pointed out that filtering out status information meets the advantage of pre-conditioned target computing nodes and be, permissible
The calculating node avoiding status information to be unsatisfactory for condition removes to download in advance foundation image, and then avoids failed download or over loading
Problem.
It may also be noted that can have multiple owing to status information meets pre-conditioned calculating node, therefore screen
The target computing nodes gone out can also have multiple.
S104, it is judged that whether there is described target computing nodes in the local mirror image warehouse of described target computing nodes and once opened
The used foundation image of container moved.
Concrete, management node judges whether there is its appearance once started in the local mirror image warehouse of target computing nodes
The used foundation image of device, if it is determined that be yes, then terminates, if it is determined that be no, then performs step S105.
Such as: assume that the container that target computing nodes once started used foundation image x, foundation image y and basis mirror
Foundation image w is only existed, then perform step S105 in z, and the local mirror image warehouse of current goal calculating node.
It is pointed out that and judge whether the used foundation image of container that target computing nodes once started
Advantage is, can avoid calculating node and download the foundation image repeated.
S105, it is judged that described target computing nodes is the most downloading foundation image.
Concrete, management node judges that target computing nodes is the most downloading foundation image, if it is determined that be yes, then ties
Bundle, if it is determined that be no, then performs step S106.
It is pointed out that the advantage judging whether to download foundation image is, can avoid calculating node and download
The foundation image repeated.
S106, to described target computing nodes send pre-download order, described pre-download order include described in once started
The mark of the used foundation image of container and the ranking results of described foundation image.
Concrete, management node is connected to target computing nodes transmission pre-download order by the communication set up, and this is pre-
Download command includes the mark of the used foundation image of container and the ranking results of foundation image once started, and target calculates
Node, after receiving pre-download order, once started according to the progressive download of ranking results from far-end mirror image warehouse in advance
The used foundation image of container.
Container cluster management framework in the embodiment of the present invention include manage node, far-end mirror image warehouse and at least one
Calculating node, management node first obtains each and calculates status information of node and each calculates making of container that node once started
By information, use information includes the mark of the used foundation image of container once started, then from all of calculating node
In filter out its status information and meet pre-conditioned target computing nodes, then send pre-download life to target computing nodes
Order, pre-download order is used for indicating target computing nodes to download foundation image in advance from far-end mirror image warehouse, in order to target calculates
The foundation image that the follow-up direct use of node is downloaded in advance starts container, thus realizes accelerating the startup time of container.
Fig. 2 is the schematic flow sheet of the startup method of a kind of container in the embodiment of the present invention, and the method is implemented in sets of containers
Management node in group's management framework.The flow process of the startup method of the container in the present embodiment may include that as shown in the figure
S201, knows container to be launched and the specification information of described container.
Concrete, management node obtains user or the container enabled instruction of miscellaneous equipment input, this container enabled instruction bag
Include the specification information of container to be launched and this container.
S202, searches the calculating node that its status information meets the service condition of described specification information requirement.
Wherein, the service condition that described specification information requires can carry out table by the utilization rate of processor and memory usage
Levying, utilization rate and the memory usage of processor are the biggest, and service condition is the best.It is to say, status information can include calculating
The processor utilization rate of node and memory usage.
Concrete, management its processor utilization rate of node checks is higher than the first predetermined threshold value, and its memory usage is high
Calculating node in the second predetermined threshold value.In the embodiment of the present invention, the first predetermined threshold value and the second predetermined threshold value can be by managing
Personnel preset, and its value is not especially limited here.
It is pointed out that the advantage managing the calculating node that node checks meets service condition is, can avoid not
Meet the calculating node deactivation container of service condition, strengthen feasibility.
S203, it may be judged whether the target being loaded with down foundation image can be filtered out from the described calculating node searched
Calculate node.
It should be understood that, by the embodiment described by Fig. 1, part calculates node and is loaded with foundation image down.Tool
Body, management node judges whether to filter out the target calculating being loaded with down foundation image from the calculating node searched
Node, if it is determined that be yes, then performs step S205, if it is determined that be no, then performs step S204.
Optionally, before step S203, the status information that management node first can calculate node according to each, determine each
The surplus resources of individual calculating node, such as: assume that in the status information of calculating node, " processor utilization rate is 30%, and internal memory makes
It is 20% by rate ", then the surplus resources that may determine that this technology node is " processor idleness 70%, internal memory idleness
80% ".Accordingly, the concrete operations of management node screening target computing nodes can be: screens from the calculating node searched
Go out the target computing nodes being loaded with down foundation image that its surplus resources is the highest.
It is pointed out that management node filters out the advantage of the highest target computing nodes of surplus resources and is, permissible
The calculating node preferentially allowing surplus resources the highest starts container, improves the utilization ratio calculating node.
S204, filters out the highest calculating node of surplus resources from the described calculating node searched and calculates joint as target
Point.
Wherein, the introduction of surplus resources can not repeat refering to the content in step S203.Concrete, management
Node filters out the highest calculating node of surplus resources as target computing nodes from calculating node of searching.
S205, sends container to described target computing nodes and starts order.
Concrete, management node sends container by the communication connection set up to target computing nodes and starts order, mesh
Mark calculating node, after receiving this container and starting order, uses the foundation image downloaded the most in advance to start container.Wherein, make
Foundation image be the foundation image that this container is mated, certainly, if the foundation image that this container is mated is not by target
Calculate node to download in advance, the most also need to download its foundation image mated from far-end mirror image warehouse.
Container cluster management framework in the embodiment of the present invention include manage node, far-end mirror image warehouse and at least one
Calculating node, management node first knows the specification information of container to be launched and this container, then searches the satisfied rule of its status information
The calculating node of the service condition of lattice information requirements, then filters out from the calculating node searched and is loaded with down basis mirror
The target computing nodes of picture, then sends container to target computing nodes and starts order, it is possible to achieve allow and be loaded with basis down
The calculating node of mirror image starts container, thus accelerates the startup time of container.
Fig. 3 is the schematic flow sheet of the delet method of a kind of foundation image in the embodiment of the present invention, and the method is implemented in appearance
Calculating node in device cluster management framework.The flow process of the delet method of the foundation image in the present embodiment can be wrapped as shown in the figure
Include:
S301, the foundation image having not been used in the local mirror image warehouse of inquiry.
It should be understood that, by the embodiment described by Fig. 1, part calculates node and is loaded with foundation image down.Also
It should be understood that, calculate node and foundation image can be used to start container, if certain foundation image is not used by for a long time, then this base is described
The use frequency of plinth mirror image is on the low side.Concrete, calculating node is at interval of a preset time period, in the local mirror image warehouse of inquiry not
Used foundation image.Wherein, preset time period can be preset by management personnel, and its value the most specifically limits
Fixed.
According to the status information of self, S302, judges whether self is in higher load condition.
Optionally, described status information includes processor utilization rate, memory usage and local mirror image warehouse utilization rate.
Concrete, calculate node and judge that the concrete operations self whether being in higher load condition are according to the status information of self: judge
Whether processor utilization rate is higher than the first predetermined threshold value, and whether memory usage is higher than the second predetermined threshold value, and local mirror image storehouse
Whether storehouse utilization rate is higher than the 3rd predetermined threshold value.In the embodiment of the present invention, the first predetermined threshold value, the second predetermined threshold value and the 3rd
Predetermined threshold value can be preset by management personnel, and its value is not especially limited here.Such as: first can be set and preset threshold
Value, the second predetermined threshold value and the 3rd predetermined threshold value are respectively 70%, 70% and 90%.
S303, the most then delete the foundation image having not been used in described local mirror image warehouse.
Concrete, calculating node if it is determined that self is in higher load condition, is then deleted in local mirror image warehouse and is not used by
The foundation image crossed.
Optionally, before step S303, calculate the idle duration of the foundation image that querying node has not been used, and press
According to the idle duration order from long to short of foundation image, foundation image is ranked up.Accordingly, knot removal is calculated local
The concrete operations of the foundation image having not been used in mirror image warehouse are: delete local mirror image successively according to the order of ranking results
The foundation image having not been used in warehouse, until self being not at higher load condition.
It is pointed out that the advantage that calculating node deletes foundation image successively according to the order of ranking results is, can
Leave unused the longer foundation image of duration with preferential deletion.
Container cluster management framework in the embodiment of the present invention include manage node, far-end mirror image warehouse and at least one
Calculate node, calculate node and first inquire about the foundation image having not been used in local mirror image warehouse, then according to the state of self
Information judges whether self is in higher load condition, then deletes when being judged to be in described local mirror image warehouse and is not used by
The foundation image crossed, can allow and be in the calculating knot removal foundation image of higher load condition, to reduce load.
Fig. 4 is a kind of structural representation managing node in the embodiment of the present invention, and this management node is implemented in container cluster
Management framework, container cluster management framework also includes far-end mirror image warehouse and at least one calculating node.The present invention as shown in the figure
Management node in embodiment at least can include data obtaining module 410, node screening module 420 and order sending module
430, wherein:
Data obtaining module 410, for obtaining status information and each described calculating node of each described calculating node
The use information of the container once started, described use information include described in the used foundation image of container that once started.
Node screening module 420, meets default bar for filtering out its status information from all of described calculating node
The target computing nodes of part.
Order sending module 430, for sending pre-download order, described pre-download order bag to described target computing nodes
Including the mark of the used foundation image of container that described target computing nodes once started, described pre-download order is used for indicating
Described target computing nodes downloads the container use that described target computing nodes once started in advance from described far-end mirror image warehouse
The foundation image crossed.
Optionally, described use information also include described in once started the used foundation image of container use frequency
Rate.Accordingly, referring to Fig. 4, the management node in the embodiment of the present invention can also include mirror image order module as shown in the figure
440, for the order from high to low of the use frequency according to described foundation image, calculating node each described was once started
The used foundation image of container is ranked up.
Described pre-download order also include described in the ranking results of the used foundation image of container that once started, described
Pre-download order is for indicating described target computing nodes suitable according to described ranking results from described far-end mirror image warehouse in advance
Once the used foundation image of container started described in sequence download.
Optional again, described status information includes the processor utilization rate of described calculating node, memory usage and basis
Mirror image warehouse, ground utilization rate.Accordingly, described node screening module 420, specifically for from all of described calculating node, sieve
Selecting its processor utilization rate and be less than the first predetermined threshold value, its memory usage is less than the second predetermined threshold value, and its local mirror
As warehouse utilization rate is less than the target computing nodes of the 3rd predetermined threshold value.
Optional again, refer to Fig. 4, the management node in the embodiment of the present invention can also include that mirror image determines as shown in the figure
, in the local mirror image warehouse determining described target computing nodes, there is not described target computing nodes once started in module 450
The used foundation image of container crossed.
Optional again, mirror image determines module 450, it is also possible to be used for determining that described target computing nodes is the most downloading base
Plinth mirror image.
Fig. 5 is the structural representation of another kind of management node in the embodiment of the present invention, and this management node is implemented in sets of containers
Group's management framework, container cluster management framework also includes far-end mirror image warehouse and at least one calculating node.As shown in the figure this
Management node in bright embodiment at least may include that
Information knows module 510, for knowing container to be launched and the specification information of described container.
Node checks module 520, meets by searching its status information based on the service condition of described specification information requirement
Operator node.
Node screening module 530, is loaded with foundation image down for filtering out from the described calculating node searched
Target computing nodes.
Order sending module 540, starts order for sending container to described target computing nodes, and described container starts life
Order starts described container for indicating the foundation image downloaded the most in advance described in the use of described target computing nodes.
Optionally, described status information includes processor utilization rate and the memory usage of described calculating node.Accordingly,
Described node checks module 520, is higher than the first predetermined threshold value specifically for searching its processor utilization rate, and its internal memory uses
Rate is higher than the calculating node of the second predetermined threshold value.
Optional again, refer to Fig. 5, the management node in the embodiment of the present invention can also include that resource determines as shown in the figure
Module 550, for according to the status information of calculating node each described, determines the surplus resources of each described calculating node.Phase
Answer, described node screening module 530, the highest specifically for filtering out its surplus resources from the described calculating node searched
It is loaded with down the target computing nodes of foundation image.
Optional again, described node screening module 530, if under being additionally operable in the described calculating node searched the most in advance
It is loaded with the target computing nodes of foundation image, the most therefrom filters out the highest calculating node of its surplus resources and calculate joint as target
Point.
Fig. 6 is a kind of structural representation calculating node in the embodiment of the present invention, and this calculating node is implemented in container cluster
Management framework, container cluster management framework also includes managing node and far-end mirror image warehouse.As shown in the figure in the embodiment of the present invention
Management node at least may include that
Mirror image enquiry module 610, for inquiring about the foundation image having not been used in local mirror image warehouse.
According to the status information of self, condition judgment module 620, for judging whether self is in higher load condition.
Mirror image removing module 630, for the most then deleting the basic mirror having not been used in described local mirror image warehouse
Picture.
Optionally, described status information includes processor utilization rate, memory usage and local mirror image warehouse utilization rate.
Accordingly, whether described condition judgment module 620, specifically for decision processor utilization rate higher than the first predetermined threshold value, internal memory
Whether utilization rate is higher than the second predetermined threshold value, and whether local mirror image warehouse utilization rate is higher than the 3rd predetermined threshold value.
Optional again, refer to Fig. 6, the calculating node in the embodiment of the present invention can also include that duration is inquired about as shown in the figure
Module 640 and mirror image order module 650, wherein:
Duration enquiry module 640, is used for the idle duration of the foundation image having not been used described in inquiring about.
Mirror image order module 650, for the order from long to short of the idle duration according to described foundation image, to described base
Plinth mirror image is ranked up.
Accordingly, described mirror image removing module 630, specifically for deleting described this locality successively according to the order of ranking results
The foundation image having not been used in mirror image warehouse, until self being not at higher load condition.
Fig. 7 is the structural representation of another the management node in the embodiment of the present invention, as it is shown in fig. 7, this management node
May include that at least one processor 701, such as CPU, at least one communication bus 702, at least one communication interface 703, deposit
Reservoir 704.Wherein, communication bus 702 is for realizing the connection communication between these assemblies;Communication interface 703, for and other
Node device carries out communication;Memorizer 704 can be high-speed RAM memorizer, it is also possible to be non-volatile memorizer (non-
Volatile memory), for example, at least one disk memory.Optionally, memorizer 704 can also is that at least one is positioned at
Away from the storage device of aforementioned processor 701, memorizer 704 stores batch processing code.Processor 701 is used for calling storage
The program code of storage in device 704, the following operation of execution:
Obtain the use information of the container that each status information calculating node once started with each calculating node, described
Use information include described in the used foundation image of container that once started;
From all of described calculating node, filter out its status information meet pre-conditioned target computing nodes;
Sending pre-download order to described target computing nodes, described pre-download order included described target computing nodes once
The mark of the used foundation image of container started, described pre-download order is used for indicating described target computing nodes in advance
The used foundation image of container that described target computing nodes once started is downloaded from described far-end mirror image warehouse.
Optionally, described use information also include described in once started the used foundation image of container use frequency
Rate.Accordingly, processor 701 filters out operation conditions from all of described calculating node and meets pre-conditioned target and calculate
Before node, also perform:
According to the use frequency order from high to low of described foundation image, calculating node each described was once started
The used foundation image of container is ranked up;
Described pre-download order also include described in the ranking results of the used foundation image of container that once started, described
Pre-download order is for indicating described target computing nodes suitable according to described ranking results from described far-end mirror image warehouse in advance
Once the used foundation image of container started described in sequence download.
Optional again, described status information includes the processor utilization rate of described calculating node, memory usage and basis
Mirror image warehouse, ground utilization rate.Accordingly, processor 701 filters out its status information from all of described calculating node and meets pre-
If the concrete operations of the target computing nodes of condition are:
From all of described calculating node, filtering out its processor utilization rate and be less than the first predetermined threshold value, its internal memory makes
By rate less than the second predetermined threshold value, and its local mirror image warehouse utilization rate is less than the target computing nodes of the 3rd predetermined threshold value.
Optional again, before processor 701 sends pre-download order to described target computing nodes, also perform:
Determine in the local mirror image warehouse of described target computing nodes and do not have what described target computing nodes once started
The used foundation image of container.
Optional again, before processor 701 sends pre-download order to described target computing nodes, also perform:
Determine that described target computing nodes is the most downloading foundation image.
Or, processor 701 operation below calling the program code of storage in memorizer 704, performing:
Know container to be launched and the specification information of described container;
Search the calculating node that its status information meets the service condition of described specification information requirement;
The target computing nodes being loaded with down foundation image is filtered out from the described calculating node searched;
Sending container to described target computing nodes and start order, described container starts order by indicating based on described target
The foundation image downloaded the most in advance described in operator node use starts described container.
Optionally, described status information includes processor utilization rate and the memory usage of described calculating node.Accordingly,
The concrete operations calculating node of the service condition that its status information meets described specification information requirement searched by processor 701:
Search its processor utilization rate and be higher than the first predetermined threshold value, and its memory usage is higher than the second predetermined threshold value
Calculate node.
Optional again, processor 701 filters out from the described calculating node searched and is loaded with foundation image down
Before target computing nodes, also perform:
According to the status information of calculating node each described, determine the surplus resources of each described calculating node.
Accordingly, processor 701 filters out the mesh being loaded with down foundation image from the described calculating node searched
Mark calculates the concrete operations of node:
Filter out, from the described calculating node searched, the mesh being loaded with down foundation image that its surplus resources is the highest
Mark calculates node.
Optional again, processor 701, according to the status information of calculating node each described, determines each described calculating node
Surplus resources after, also perform:
If the described calculating node searched does not has to be loaded with down the target computing nodes of foundation image, the most therefrom sieve
Select the highest calculating node of its surplus resources as target computing nodes.
Fig. 8 is the another kind of structural representation calculating node in the embodiment of the present invention, as shown in Figure 8, this calculating node
May include that at least one processor 801, such as CPU, at least one communication bus 802, at least one communication interface 803, deposit
Reservoir 804.Wherein, communication bus 802 is for realizing the connection communication between these assemblies;Communication interface 803, for and other
Node device carries out communication;Memorizer 804 can be high-speed RAM memorizer, it is also possible to be non-volatile memorizer (non-
Volatile memory), for example, at least one disk memory.Optionally, memorizer 804 can also is that at least one is positioned at
Away from the storage device of aforementioned processor 801, memorizer 804 stores batch processing code.Processor 801 is used for calling storage
The program code of storage in device 804, the following operation of execution:
The foundation image having not been used in the local mirror image warehouse of inquiry;
Status information according to self judges whether self is in higher load condition;
The most then delete the foundation image having not been used in described local mirror image warehouse.
Optionally, described status information includes processor utilization rate, memory usage and local mirror image warehouse utilization rate.
Accordingly, according to the status information of self, processor 801 judges that the concrete operations self whether being in higher load condition are:
Whether decision processor utilization rate is higher than the first predetermined threshold value, and whether memory usage is higher than the second predetermined threshold value,
And whether local mirror image warehouse utilization rate is higher than the 3rd predetermined threshold value.
Optional again, before the foundation image having not been used in described local mirror image warehouse deleted by processor 801, also hold
OK:
The idle duration of the foundation image having not been used described in inquiry;
According to the idle duration order from long to short of described foundation image, described foundation image is ranked up.
Accordingly, the concrete operations of the foundation image that processor 801 has not been used in deleting described local mirror image warehouse
For:
The foundation image having not been used in described local mirror image warehouse is deleted successively according to the order of ranking results, until
Till self is not at higher load condition.
Container cluster management framework in the embodiment of the present invention include manage node, far-end mirror image warehouse and at least one
Calculating node, management node first obtains each and calculates status information of node and each calculates making of container that node once started
By information, use information includes the used foundation image of container once started, then screens from all of calculating node
Go out its status information and meet pre-conditioned target computing nodes, then send pre-download order to target computing nodes, pre-under
Carry order to be used for indicating target computing nodes to download foundation image in advance from far-end mirror image warehouse, in order to target computing nodes is follow-up
The foundation image that directly use is downloaded in advance starts container, thus realizes accelerating the startup time of container.
One of ordinary skill in the art will appreciate that all or part of flow process realizing in above-described embodiment method, be permissible
Instructing relevant hardware by computer program to complete, described program can be stored in a computer read/write memory medium
In, this program is upon execution, it may include such as the flow process of the embodiment of above-mentioned each method.Wherein, described storage medium can be magnetic
Dish, CD, read-only store-memory body (Read-Only Memory, ROM) or random store-memory body (Random Access
Memory, RAM) etc..
The above disclosed present pre-ferred embodiments that is only, can not limit the right model of the present invention with this certainly
Enclose, the equivalent variations therefore made according to the claims in the present invention, still belong to the scope that the present invention is contained.
Claims (24)
1. the method for down loading of a foundation image, it is characterised in that described method is implemented in the pipe in container cluster management framework
Reason node, described container cluster management framework also includes far-end mirror image warehouse and at least one calculating node, and described method includes:
Obtain the status information of each described calculating node and the use information of container that each described calculating node once started,
Described use information include described in the used foundation image of container that once started;
From all of described calculating node, filter out its status information meet pre-conditioned target computing nodes;
Send pre-download order to described target computing nodes, described pre-download order include described in the container that once started use
The mark of the foundation image crossed, described pre-download order is used for indicating described target computing nodes in advance from described far-end mirror image storehouse
Once the used foundation image of container started described in the download of storehouse.
2. the method for claim 1, it is characterised in that described use information also include described in the container that once started make
The use frequency of used foundation image;
Described filter out before operation conditions meets pre-conditioned target computing nodes from all of described calculating node, also
Including:
According to the use frequency order from high to low of described foundation image, the container that calculating node each described was once started
Used foundation image is ranked up;
Described pre-download order also include described in the ranking results of the used foundation image of container that once started, described pre-under
Carry order for indicating described target computing nodes in advance from the described far-end mirror image warehouse order according to described ranking results
Once the used foundation image of container started described in load.
3. the method for claim 1, it is characterised in that described status information includes that the processor of described calculating node makes
By rate, memory usage and local mirror image warehouse utilization rate;
Described from all of described calculating node, filter out its status information meet pre-conditioned target computing nodes, bag
Include:
From all of described calculating node, filter out its processor utilization rate less than the first predetermined threshold value, its memory usage
Less than the second predetermined threshold value, and its local mirror image warehouse utilization rate is less than the target computing nodes of the 3rd predetermined threshold value.
4. the method as described in any one of claim 1-3, it is characterised in that described under the transmission of described target computing nodes is pre-
Before carrying order, also include:
Determine in the local mirror image warehouse of described target computing nodes and there is not the container that described target computing nodes once started
Used foundation image.
5. the method as described in any one of claim 1-3, it is characterised in that described under the transmission of described target computing nodes is pre-
Before carrying order, also include:
Determine that described target computing nodes is the most downloading foundation image.
6. the startup method of a container, it is characterised in that the management joint that described method is implemented in container cluster management framework
Point, described container cluster management framework also includes far-end mirror image warehouse and at least one calculating node, and described method includes:
Know container to be launched and the specification information of described container;
Search the calculating node that its status information meets the service condition of described specification information requirement;
The target computing nodes being loaded with down foundation image is filtered out from the described calculating node searched;
Sending container to described target computing nodes and start order, described container starts order and is used for indicating described target to calculate joint
The foundation image downloaded the most in advance described in some use starts described container.
7. method as claimed in claim 6, it is characterised in that described status information includes that the processor of described calculating node makes
With rate and memory usage;
Its status information of described lookup meets the calculating node of the service condition of described specification information requirement, including:
Search its processor utilization rate and be higher than the first predetermined threshold value, and its memory usage is higher than the calculating of the second predetermined threshold value
Node.
8. method as claimed in claim 6, it is characterised in that described filter out the most in advance from the described calculating node searched
Under be loaded with foundation image target computing nodes before, also include:
According to the status information of calculating node each described, determine the surplus resources of each described calculating node;
Described filtering out from the described calculating node searched is loaded with down the target computing nodes of foundation image, including:
Filter out, from the described calculating node searched, the target meter being loaded with down foundation image that its surplus resources is the highest
Operator node.
9. method as claimed in claim 8, it is characterised in that described according to the status information of calculating node each described, really
After the surplus resources of each described calculating node fixed, also include:
If the described calculating node searched does not has to be loaded with down the target computing nodes of foundation image, the most therefrom filter out
The highest calculating node of its surplus resources is as target computing nodes.
10. the delet method of a foundation image, it is characterised in that described method is implemented in the meter in container cluster management framework
Operator node, described container cluster management framework also includes far-end mirror image warehouse and management node, and described method includes:
The foundation image having not been used in the local mirror image warehouse of inquiry;
Status information according to self judges whether self is in higher load condition;
The most then delete the foundation image having not been used in described local mirror image warehouse.
11. methods as claimed in claim 10, it is characterised in that described status information includes that processor utilization rate, internal memory make
By rate and local mirror image warehouse utilization rate;
The described status information according to self judges whether self is in higher load condition, including:
Whether decision processor utilization rate is higher than the first predetermined threshold value, and whether memory usage is higher than the second predetermined threshold value, and this
Whether mirror image warehouse, ground utilization rate is higher than the 3rd predetermined threshold value.
12. methods as claimed in claim 10, it is characterised in that described deletion has not been used in described local mirror image warehouse
Foundation image before, also include:
The idle duration of the foundation image having not been used described in inquiry;
According to the idle duration order from long to short of described foundation image, described foundation image is ranked up;
The foundation image having not been used in the described local mirror image warehouse of described deletion, including:
The foundation image having not been used in described local mirror image warehouse is deleted successively according to the order of ranking results, until self
Till being not at higher load condition.
13. 1 kinds of management nodes, it is characterised in that described management node is implemented in container cluster management framework, described sets of containers
Group's management framework also includes far-end mirror image warehouse and at least one calculating node, and described management node includes:
Data obtaining module, once started for the status information and each described calculating node obtaining each described calculating node
The use information of container, described use information include described in the used foundation image of container that once started;
Node screening module, meets pre-conditioned target for filtering out its status information from all of described calculating node
Calculate node;
Order sending module, for sending pre-download order to described target computing nodes, described pre-download order includes described
The mark of the used foundation image of container once started, described pre-download order is used for indicating described target computing nodes pre-
First from the used foundation image of container once started described in the download of described far-end mirror image warehouse.
14. manage node as claimed in claim 13, it is characterised in that described use information also include described in once started
The use frequency of the used foundation image of container;
Described management node also includes:
Mirror image order module, for the order from high to low of the use frequency according to described foundation image, to calculating each described
The used foundation image of container that node once started is ranked up;
Described pre-download order also include described in the ranking results of the used foundation image of container that once started, described pre-under
Carry order for indicating described target computing nodes in advance from the described far-end mirror image warehouse order according to described ranking results
Once the used foundation image of container started described in load.
15. manage node as claimed in claim 13, it is characterised in that described status information includes the place of described calculating node
Reason device utilization rate, memory usage and local mirror image warehouse utilization rate;
Described node screening module, specifically for from all of described calculating node, filters out its processor utilization rate and is less than
First predetermined threshold value, its memory usage is less than the second predetermined threshold value, and its local mirror image warehouse utilization rate is pre-less than the 3rd
If the target computing nodes of threshold value.
The 16. management nodes as described in any one of claim 13-15, it is characterised in that described management node also includes:
Mirror image determines module, there is not described target and calculate joint in the local mirror image warehouse determining described target computing nodes
The used foundation image of container that point once started.
The 17. management nodes as described in any one of claim 13-15, it is characterised in that described management node also includes:
Mirror image determines module, is used for determining that described target computing nodes is the most downloading foundation image.
18. 1 kinds of management nodes, it is characterised in that described management node is implemented in container cluster management framework, described sets of containers
Group's management framework also includes far-end mirror image warehouse and at least one calculating node, and described management node includes:
Information knows module, for knowing container to be launched and the specification information of described container;
Node checks module, for searching the calculating node that its status information meets the service condition of described specification information requirement;
Node screening module, is loaded with based on the target of foundation image down by filtering out from the described calculating node searched
Operator node;
Order sending module, starts order for sending container to described target computing nodes, and described container starts order and is used for
The foundation image downloaded the most in advance described in the use of described target computing nodes is indicated to start described container.
19. manage node as claimed in claim 18, it is characterised in that described status information includes the place of described calculating node
Reason device utilization rate and memory usage;
Described node checks module, is higher than the first predetermined threshold value specifically for searching its processor utilization rate, and its internal memory makes
The calculating node of the second predetermined threshold value it is higher than by rate.
20. manage node as claimed in claim 18, it is characterised in that described management node also includes:
Resource determination module, for according to the status information of calculating node each described, determining the surplus of each described calculating node
Remaining resource;
Described node screening module, specifically for filtering out the highest the most pre-of its surplus resources from the described calculating node searched
It is loaded with down the target computing nodes of foundation image.
21. manage node as claimed in claim 20, it is characterised in that described node screening module, if being additionally operable to lookup
Described calculating node does not has to be loaded with down the target computing nodes of foundation image, the most therefrom filters out its surplus resources
High calculating node is as target computing nodes.
22. 1 kinds calculate node, it is characterised in that described calculating node is implemented in container cluster management framework, described sets of containers
Group's management framework also includes far-end mirror image warehouse and management node, and described method includes:
Mirror image enquiry module, for inquiring about the foundation image having not been used in local mirror image warehouse;
According to the status information of self, condition judgment module, for judging whether self is in higher load condition;
Mirror image removing module, for the most then deleting the foundation image having not been used in described local mirror image warehouse.
23. calculate node as claimed in claim 22, it is characterised in that described status information includes processor utilization rate, interior
Deposit utilization rate and local mirror image warehouse utilization rate;
Whether described condition judgment module, specifically for decision processor utilization rate higher than the first predetermined threshold value, memory usage
Whether higher than the second predetermined threshold value, and whether local mirror image warehouse utilization rate is higher than the 3rd predetermined threshold value.
24. calculate node as claimed in claim 22, it is characterised in that described calculating node also includes:
Duration enquiry module, is used for the idle duration of the foundation image having not been used described in inquiring about;
Mirror image order module, for the order from long to short of the idle duration according to described foundation image, to described foundation image
It is ranked up;
Described mirror image removing module, specifically for according to the order of ranking results delete successively in described local mirror image warehouse not by
Used foundation image, until self being not at higher load condition.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610586890.8A CN106302632B (en) | 2016-07-21 | 2016-07-21 | Downloading method of basic mirror image and management node |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610586890.8A CN106302632B (en) | 2016-07-21 | 2016-07-21 | Downloading method of basic mirror image and management node |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106302632A true CN106302632A (en) | 2017-01-04 |
CN106302632B CN106302632B (en) | 2020-02-14 |
Family
ID=57652226
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610586890.8A Active CN106302632B (en) | 2016-07-21 | 2016-07-21 | Downloading method of basic mirror image and management node |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106302632B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106911777A (en) * | 2017-02-24 | 2017-06-30 | 郑州云海信息技术有限公司 | A kind of data processing method and server |
CN107493191A (en) * | 2017-08-08 | 2017-12-19 | 深信服科技股份有限公司 | A kind of clustered node and self scheduling container group system |
CN109062658A (en) * | 2018-06-29 | 2018-12-21 | 优刻得科技股份有限公司 | Realize dispatching method, device, medium, equipment and the system of computing resource serviceization |
CN109104451A (en) * | 2017-06-21 | 2018-12-28 | 阿里巴巴集团控股有限公司 | The pre-heating mean and node of the method for down loading and node of Docker mirror image, Docker mirror image |
CN109343859A (en) * | 2018-08-17 | 2019-02-15 | 咪咕文化科技有限公司 | A kind of information processing method, device and storage medium |
CN109413069A (en) * | 2018-10-29 | 2019-03-01 | 北京百悟科技有限公司 | The application method and device of virtual site firewall based on block chain |
CN109947438A (en) * | 2018-12-28 | 2019-06-28 | 广州卓动信息科技有限公司 | Game installation method and electronic equipment, readable storage medium storing program for executing based on machine learning |
CN110008006A (en) * | 2019-04-11 | 2019-07-12 | 中国联合网络通信集团有限公司 | Big data tool dispositions method and system based on container |
CN110287004A (en) * | 2019-07-05 | 2019-09-27 | 中国工商银行股份有限公司 | Basic environment mirror image pre-heating mean and device based on docker container technique |
CN111078402A (en) * | 2019-12-03 | 2020-04-28 | 广州西麦科技股份有限公司 | Resource pool system capable of rapidly providing experimental environment |
CN111290776A (en) * | 2018-12-10 | 2020-06-16 | 航天信息股份有限公司 | Management device and method for managing block chain nodes |
CN117850980A (en) * | 2023-12-25 | 2024-04-09 | 慧之安信息技术股份有限公司 | Container mirror image construction method and system based on cloud edge cooperation |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102932408A (en) * | 2012-09-21 | 2013-02-13 | 中国电信股份有限公司云计算分公司 | Management method and platform of server cluster |
CN105227374A (en) * | 2015-10-23 | 2016-01-06 | 浪潮(北京)电子信息产业有限公司 | A kind of fault moving method of cluster application and system |
CN105468362A (en) * | 2015-11-17 | 2016-04-06 | 广州杰赛科技股份有限公司 | Application deployment method and cloud computing system |
CN105677479A (en) * | 2015-12-30 | 2016-06-15 | 北京奇艺世纪科技有限公司 | Method and device for implementing parallel running of GPU operation programs |
CN105763545A (en) * | 2016-02-04 | 2016-07-13 | 杭州华三通信技术有限公司 | BYOD method and device |
-
2016
- 2016-07-21 CN CN201610586890.8A patent/CN106302632B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102932408A (en) * | 2012-09-21 | 2013-02-13 | 中国电信股份有限公司云计算分公司 | Management method and platform of server cluster |
CN105227374A (en) * | 2015-10-23 | 2016-01-06 | 浪潮(北京)电子信息产业有限公司 | A kind of fault moving method of cluster application and system |
CN105468362A (en) * | 2015-11-17 | 2016-04-06 | 广州杰赛科技股份有限公司 | Application deployment method and cloud computing system |
CN105677479A (en) * | 2015-12-30 | 2016-06-15 | 北京奇艺世纪科技有限公司 | Method and device for implementing parallel running of GPU operation programs |
CN105763545A (en) * | 2016-02-04 | 2016-07-13 | 杭州华三通信技术有限公司 | BYOD method and device |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106911777A (en) * | 2017-02-24 | 2017-06-30 | 郑州云海信息技术有限公司 | A kind of data processing method and server |
CN109104451A (en) * | 2017-06-21 | 2018-12-28 | 阿里巴巴集团控股有限公司 | The pre-heating mean and node of the method for down loading and node of Docker mirror image, Docker mirror image |
CN107493191B (en) * | 2017-08-08 | 2020-12-22 | 深信服科技股份有限公司 | Cluster node and self-scheduling container cluster system |
CN107493191A (en) * | 2017-08-08 | 2017-12-19 | 深信服科技股份有限公司 | A kind of clustered node and self scheduling container group system |
CN109062658A (en) * | 2018-06-29 | 2018-12-21 | 优刻得科技股份有限公司 | Realize dispatching method, device, medium, equipment and the system of computing resource serviceization |
CN109343859A (en) * | 2018-08-17 | 2019-02-15 | 咪咕文化科技有限公司 | A kind of information processing method, device and storage medium |
CN109413069A (en) * | 2018-10-29 | 2019-03-01 | 北京百悟科技有限公司 | The application method and device of virtual site firewall based on block chain |
CN109413069B (en) * | 2018-10-29 | 2021-11-12 | 北京百悟科技有限公司 | Application method and device of virtual website firewall based on block chain |
CN111290776B (en) * | 2018-12-10 | 2023-05-16 | 航天信息股份有限公司 | Management device and method for managing blockchain nodes |
CN111290776A (en) * | 2018-12-10 | 2020-06-16 | 航天信息股份有限公司 | Management device and method for managing block chain nodes |
CN109947438A (en) * | 2018-12-28 | 2019-06-28 | 广州卓动信息科技有限公司 | Game installation method and electronic equipment, readable storage medium storing program for executing based on machine learning |
CN110008006B (en) * | 2019-04-11 | 2021-04-02 | 中国联合网络通信集团有限公司 | Container-based big data tool deployment method and system |
CN110008006A (en) * | 2019-04-11 | 2019-07-12 | 中国联合网络通信集团有限公司 | Big data tool dispositions method and system based on container |
CN110287004B (en) * | 2019-07-05 | 2021-07-30 | 中国工商银行股份有限公司 | Basic environment mirror image preheating method and device based on docker container technology |
CN110287004A (en) * | 2019-07-05 | 2019-09-27 | 中国工商银行股份有限公司 | Basic environment mirror image pre-heating mean and device based on docker container technique |
CN111078402A (en) * | 2019-12-03 | 2020-04-28 | 广州西麦科技股份有限公司 | Resource pool system capable of rapidly providing experimental environment |
CN117850980A (en) * | 2023-12-25 | 2024-04-09 | 慧之安信息技术股份有限公司 | Container mirror image construction method and system based on cloud edge cooperation |
Also Published As
Publication number | Publication date |
---|---|
CN106302632B (en) | 2020-02-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106302632A (en) | The method for down loading of a kind of foundation image and management node | |
CN106201661B (en) | Method and apparatus for elastic telescopic cluster virtual machine | |
CN104778178B (en) | A kind of application class method, apparatus and service server | |
CN107977252A (en) | A kind of capacity reduction method, device and the cloud platform of cloud platform business | |
CN104899049A (en) | File deletion method and device | |
CN102508693A (en) | Web server capacity expansion system based on virtual machine | |
CN108595503A (en) | Document handling method and server | |
CN104657380A (en) | Picture loading method and device | |
CN109857548A (en) | Memory space optimization method, device, computer equipment and storage medium | |
CN103412825B (en) | A kind of memory release method and device | |
CN108762792A (en) | Upgrading network equipment method, upgrade server, terminal device and storage medium | |
CN108121511A (en) | Data processing method, device and equipment in a kind of distributed edge storage system | |
CN107656776A (en) | A kind of firmware update, device and the medium of FPGA accelerator cards | |
CN103500191A (en) | Flow table configuration, query and table item deleting method and device | |
CN112114849B (en) | Resource updating method and device | |
CN103761136A (en) | Method for dynamically loading data model based on plug-in | |
CN103473343A (en) | File management method, device and terminal | |
CN101246440B (en) | Java card system rubbish recovering method | |
CN106331160A (en) | Data migration method and system | |
CN108984295A (en) | Method for recovering internal storage, computer installation and computer readable storage medium | |
CN106201561A (en) | The upgrade method of distributed caching cluster and equipment | |
CN107800814A (en) | Virtual machine deployment method and device | |
CN103179153B (en) | Mobile terminal and mobile terminal data backup method based on cloud server | |
CN107403012A (en) | A kind of method for interchanging data and device | |
CN113721936B (en) | Application management method, intelligent terminal, device and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |