CN108628613A - The implementation method of the stateful service of container cluster based on domestic CPU and OS - Google Patents

The implementation method of the stateful service of container cluster based on domestic CPU and OS Download PDF

Info

Publication number
CN108628613A
CN108628613A CN201810411194.2A CN201810411194A CN108628613A CN 108628613 A CN108628613 A CN 108628613A CN 201810411194 A CN201810411194 A CN 201810411194A CN 108628613 A CN108628613 A CN 108628613A
Authority
CN
China
Prior art keywords
container group
stateful
container
service
state
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.)
Pending
Application number
CN201810411194.2A
Other languages
Chinese (zh)
Inventor
朱宪
许蕾
李超
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong Hui Trade Electronic Port Co Ltd
Shandong Huimao Electronic Port Co Ltd
Original Assignee
Shandong Hui Trade Electronic Port Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shandong Hui Trade Electronic Port Co Ltd filed Critical Shandong Hui Trade Electronic Port Co Ltd
Priority to CN201810411194.2A priority Critical patent/CN108628613A/en
Publication of CN108628613A publication Critical patent/CN108628613A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F8/00Arrangements for software engineering
    • G06F8/60Software deployment

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The present invention discloses the implementation method of the stateful service of container cluster based on domestic CPU and OS, is related to distributed container cluster administrative skill field;The particular order of initial work and control container the group deployment of stateful container in cluster is carried out by state controller component, the network identity of container group in cluster is provided using zero-address service, maintain the stable relation of cluster member, it can be used persistent storage volume as stable persistent storage, it names container group, zero-address service and persistent storage to state using serializing simultaneously, realizes resource and application logical separation and one-to-one correspondence mutually binding;Stateful service cluster is managed by stateful copy controller assemblies, the state of detection container group, so as to find operation node failure stateful service container group and carry out self-regeneration, improve the availability of container service, the orderly dilatation of stateful service cluster and orderly capacity reducing can be realized simultaneously, improve the redundancy ability of container cluster.

Description

The implementation method of the stateful service of container cluster based on domestic CPU and OS
Technical field
The present invention discloses a kind of implementation method of the stateful service of container cluster, is related to distributed container cluster administrative skill Field, the specifically implementation method of the stateful service of container cluster based on domestic CPU and OS.
Background technology
With the development of domestic enterprise's informationization and electronic government affairs, large and medium-sized enterprise and government organs increasingly according to Rely and the running of its business is unfolded in information system, software application is changed into incognito at domestic CPU and operating system by single architecture Business framework is increasingly becoming trend, using deployment also by gradual containerization, clustering.
The application disposed in the cluster, which is typically necessary, is designed as stateless service, what this kind of service need not be locally stored Perdurable data, and multiple examples are completely the same to the result of the same request response, and convenient container management platform manages it Reason, can restart or dilatation at any time, not interfere with the service of container.In addition to stateless service, user is also more and more to have use to have The demand of status service, stateful service are the services for needing preservation state and restoring state, and example is needed data at any time It is backed up, and when creating a new stateful service, it can be with these data of Backup and Restore, stateful service composition collection If group, sufficiently complex, current container cluster cannot be satisfied the demand of this stateful service.
The present invention provides the implementation method of the stateful service of container cluster based on domestic CPU and OS, in stateful service In cluster, the network address that each container is required for fixed name and stablizes is in communication with each other for cluster inner pressurd vessel, needs Respective fixed storage, is preserved and recovered respective state.The case where for above-mentioned stateful service cluster, the method for the present invention are Container service provides stable network address, stable persistent storage, orderly deployment and extension, orderly deletion and end Only, orderly automatic rolling update, the state for solving stateful service preserves and cluster problem of management, realizes stateful clothes Business, solves the deployment and management of the stateful service of the safe and reliable cloud platform of domestic environment.
Invention content
In view of the shortcomings of the prior art and problem, provide the container cluster based on domestic CPU and OS has shape to the present invention The implementation method of state service,
Concrete scheme proposed by the present invention is:
The realization system of the stateful service of container cluster based on domestic CPU and OS, including distributed container cluster management platform The state controller component of operation and stateful copy controller assemblies,
Stateful copy controller assemblies are responsible for the stateful service cluster of container cluster, the related stateful service of identification Order executes corresponding stateful service according to the operating condition state of a control controller assemblies of management platform respective resources and is closed Join the operational order of container group, while detecting the state of the container group;
State controller component is responsible for disposing the container group, is named for the container group, provides described in zero-address service conduct The network identity of container group provides persistent storage for the container group and states and bound, and starts to container group suitable Sequence is controlled, while executing specific related behaviour for the container group according to the order of stateful copy controller assemblies Make.
The order of the related stateful service of stateful copy controller assemblies identification in the system, when one it is stateful The associated container group copy amount of service is excessive, and it is extra that stateful copy controller assemblies state of a control controller assemblies are deleted Container group copy, on the contrary stateful copy controller assemblies state of a control controller assemblies create container group copy amount, until Container group copy amount associated by the stateful service reaches preset copy amount.
The stateful service cluster of stateful copy controller assemblies management container cluster, detection have shape in the system The state of the associated container group of state service, the health status of the container group is obtained every the set time, detects container group Abnormal, then stateful copy controller assemblies rebuild abnormal container group by state controller component or dispatch state is normally held Device group is replaced, and the container group replaced possesses whole resources of abnormal container group.
Container group in the system associated by the stateful service of state controller deployment of components, according to the container group Container group described in title and serial number renames, and the new name of the container group is as the unique identities mark in container cluster Know, and the order-assigned zero-address service newly named according to the container group, and binding persistent storage.
State controller component is adjusted in certain sequence according to the order of stateful copy controller assemblies in the system With the other assemblies in distributed container cluster management platform order is executed according to determining sequence.
State controller component is according to the newly-built orders of stateful copy controller assemblies in the system, for described Container group is renamed according to container group described in the container group name and serial number, and is newly named according to the container group suitable The service of ordered pair zero-address and persistent storage are named, and successively to the identical name container group of serial number, zero-address service and Persistent storage is bound.
State controller component is according to the reconstruction orders of stateful copy controller assemblies in the system, for there is shape The abnormal container group of state service, suppressing exception container group, and abnormal container group is rebuild, the container group after reconstruction uses abnormal container The name and zero-address service of group and persistent storage.
The implementation method of the stateful service of container cluster based on domestic CPU and OS, utilizes distributed container cluster management The stateful service cluster of the stateful copy controller assemblies management container cluster of platform operation, the related stateful service of identification Order, execute corresponding stateful service institute according to the operating condition state of a control controller assemblies of management platform respective resources The operational order of associated container group, while detecting the state of the container group;
Container group described in state controller deployment of components using the operation of distributed container cluster management platform is the container group Name, provides network identity of the zero-address service as the container group, and persistent storage statement is provided simultaneously for the container group It is bound, and the sequence started to container group controls, while according to the order of stateful copy controller assemblies, for The container group, executes specific relevant operation.
Container group described in state controller deployment of components in the method,
When a stateful service is only associated with a container group, specific steps:
Step 1:Check whether the container group has carried out the generation and binding of persistent storage statement, if not generating and binding, Step 1 is executed, it is no to then follow the steps 2;
Step 2:It continues checking for whether zero-address service creates, does not create, then execute the zero-address service that creates, it is no to then follow the steps 3;
Step 3:Using fixed container group name, start container group, initialization container group state;
When a stateful N number of container group of service cluster respective associated, state controller component is according to serial number from 0 to the suitable of N-1 Sequence deployment container group, the specific steps are:
Step 1:Whether the container group state for detecting serial number 0 is normal, if normally executing step 2, otherwise deployment container group -0, Then step 1 is executed;
Step 2:Container group serial number adds 1, and whether detection container group state is normal, if normal and serial number is less than N-1, continues to execute Step 2, until container group serial number N-1 then terminates;Otherwise the container group of the serial number is disposed, step 1 is then executed;Wherein N is Positive integer.
The order of the related stateful service of stateful copy controller assemblies identification in the method, when one it is stateful The associated container group copy amount of service is excessive, and it is extra that stateful copy controller assemblies state of a control controller assemblies are deleted Container group copy, on the contrary stateful copy controller assemblies state of a control controller assemblies create container group copy amount, until Container group copy amount associated by the stateful service reaches preset copy amount.
Usefulness of the present invention is:
The present invention provides the implementation methods based on domestic CPU and the distributed stateful service of container cluster of operating system, can The safe and reliable container cloud platform that whole nation production based on domestic CPU and operating system is environmentally disposed, realizes that stateful service is held The deployment operation of device group and cluster management, and pass through the initial chemical industry of stateful container in state controller component progress cluster The particular order for making simultaneously control container group deployment, the network identity of container group in cluster is provided using zero-address service, is maintained The stable relation of cluster member can be used persistent storage volume as stable persistent storage, while use serializing name Resource and application logical separation and one-to-one correspondence mutually binding are realized in container group, zero-address service and persistent storage statement;It is logical Stateful copy controller assemblies are crossed to manage stateful service cluster, the state of detection container group, to find operation node The container group of the stateful service of failure simultaneously carries out self-regeneration, improves the availability of container service, while can realize stateful The orderly dilatation of service cluster and orderly capacity reducing improve the redundancy ability of container cluster.
Description of the drawings
Fig. 1 is the stateful copy controller assemblies and state controller component relation schematic diagram of present system;
Fig. 2 is that the state controller component of present system and distributed container cluster management platform other assemblies relationship are illustrated Figure;
Fig. 3 is stateful service cluster resource binding relationship schematic diagram in the present invention;
Fig. 4 is the flow diagram for the container group that state controller deployment of components creates associated by stateful service cluster;
Fig. 5 is the flow diagram for the container group that state controller deployment of components is deleted associated by stateful service cluster;
Fig. 6 is the flow diagram that stateful copy controller assemblies state of a control controller assemblies carry out container group capacity reducing;
Fig. 7 is the flow diagram that stateful copy controller assemblies state of a control controller assemblies carry out container group dilatation.
Specific implementation mode
The present invention provides the implementation method of the stateful service of container cluster based on domestic CPU and OS, is held using distribution The stateful service cluster of the stateful copy controller assemblies management container cluster of device cluster management platform operation, identification are related The order of stateful service has accordingly according to the execution of the operating condition state of a control controller assemblies of management platform respective resources The operational order of container group associated by status service, while detecting the state of the container group;
Container group described in state controller deployment of components using the operation of distributed container cluster management platform is the container group Name, provides network identity of the zero-address service as the container group, and persistent storage statement is provided simultaneously for the container group It is bound, and the sequence started to container group controls, while according to the order of stateful copy controller assemblies, for The container group, executes specific relevant operation.
The realization system of the stateful service of container cluster based on domestic CPU and OS corresponding with the above method is provided simultaneously System, including the state controller component of distributed container cluster management platform operation and stateful copy controller assemblies,
Stateful copy controller assemblies are responsible for the stateful service cluster of container cluster, the related stateful service of identification Order executes corresponding stateful service according to the operating condition state of a control controller assemblies of management platform respective resources and is closed Join the operational order of container group, while detecting the state of the container group;
State controller component is responsible for disposing the container group, is named for the container group, provides described in zero-address service conduct The network identity of container group provides persistent storage for the container group and states and bound, and starts to container group suitable Sequence is controlled, while executing specific related behaviour for the container group according to the order of stateful copy controller assemblies Make.
In conjunction with attached drawing and specific implementation, the technical solution that further illustrates the present invention.
Stateful copy controller assemblies and state controller component can be solidificated in module in present system, so as to Hardware adaptations are carried out, wherein the major function of stateful copy controller assemblies is order analysis and the stateful associated appearance of service The function that device group is monitored, wherein being ordered accordingly container group according to distributed container cluster management platform, stateful pair This controller assemblies analyzes the execution classification of the order according to the operating condition of respective resources in container cluster, identification is newly-built, The operation behaviors such as dilatation, capacity reducing, deletion, and state of a control controller assemblies are executed, with reference to figure 1;
Stateful copy controller assemblies just obtain the state of container group every the set time simultaneously, it is ensured that stateful in cluster The associated container group of service preserves the state that a certain number of container group copies are in normal operation, if the stateful service Container group copy amount it is too many, then stateful copy controller assemblies destroy some container group copies;Otherwise stateful copy Controller assemblies add container group copy, until the quantity of the container group copy of the stateful service reaches preset number of copies Amount improves the redundancy ability of present system by stateful copy controller assemblies,
And stateful copy controller assemblies obtain the state of container group, if it find that container group is abnormal, such as container group institute In node collapses, and when obtaining the information less than container group normal operation, then confirm that container group is abnormal container group, then there is shape State copy controller assemblies pass through state controller component again scheduling and abnormal container group same state on other nodes Container group replaces, whole resources of abnormal container group before the container group of replacement possesses, including container group name, network mark Know, the persistent storage statement bound with persistent storage volume, and the effect in container cluster.
State controller component mainly has NAPT function, control function and Reconstruction of The Function etc. to container group, with reference to figure 2,
When the newly-built order for receiving stateful copy controller assemblies, state controller component first initializes container group Work, i.e., to container group, zero-address service, persistent storage volume statement carry out sequence name, and successively to identical name serial number Container group, zero-address service and persistent storage volume statement bound;Wherein the data of container group are all under default situations Non-persistentization, data are also and then lost after the extinction of container group, but Docker provides volume mechanism so as to by lasting data Change storage, in distributed container cluster of the invention, provides more powerful volume mechanism and plug-in unit, realize that the persistence of data is deposited Storage, persistent storage is one section of network storage by administrator configurations in cluster, by the volume mechanism of distributed container cluster and Memory card distributes to the use of container group, to realize the persistent storage of data, has independently of times for using persistent storage The characteristics of life cycle of what single container group;
Such as state controller component receives newly-built order, when disposing the stateful service of a N number of container group, is used according to platform Family provide container group name container group is renamed, according to " container group name-serial number " rule, name container group be " hold Device group name -0 ", " container group name -1 " and so on arrive " container group name-N-1 ", these titles are fixed, and container group Name can be as the unique identity in cluster;Order-assigned is carried out according to the zero-address service name that user provides, " container group name -0 " correspondence " zero-address service name -0 ", " container group name -1 " correspondence " zero-address service name -1 " and so on To " container group name-N-1 " correspondence " zero-address service name-N-1 ";According to user's needs, persistence is distributed for each container group " container group name -0 " is distributed in storage statement, " persistent storage statement -0 ", and " persistent storage statement -1 " distributes to " container group Name -1 ", and so on distributed to " container group name-N-1 " to " persistent storage statement-N-1 ", and before starting container group, shape State controller assemblies ensure that each persistent storage statement and persistent storage binding success, persistent storage can be carried by administrator Preceding offer can also realize dynamic provisioning with dynamic memory class, and persistent storage statement is user to persistent storage Request allows user using storage resource is abstracted, it is similar to container group, and container group consumes node resource, persistent storage sound Bright consumption persistence resource;And the container group that zero-address service is stateful service distributes DNS domain name as its network identity, Other services or container group can access stateful service and its rear end container group by DNS domain name in container cluster,
I.e. in distributed container cluster, under default situations, a virtual cluster ip is distributed for each service, agency can be passed through Mode realizes that cluster ip is accessed, and zero-address service refers to not distributing the addresses cluster ip in distributed container cluster, passes through appearance The DNS domain name service of device cluster internal realizes that zero-address accesses, when the service name in access queries container cluster, DNS meetings Multiple A records are returned to, these records are all directed to zero-address and service corresponding rear end container group address, with reference to figure 3;Wherein HSName is the first character section of DNS domain name, and DNS domain name format is:Name.namespace.cluster.local is wherein Name is the title of service, and namespace is the name space in container cluster where service, and cluster.local is container The domain name for the container cluster that cluster management platform defines, HSName is exactly the title serviced herein, under same namespace, The title is accessed to be equivalent to access HSName.namespace.cluster.local;
Stateful service related resources are generated and dispatch state controller assemblies travel control function, state controller component exists After container group carries out serializing name, in certain sequence, the other assemblies in distributed container cluster management platform are called, such as It calls persistent storage declaration management component to generate persistent storage statement, and executes and carry out the binding of persistent storage volume, It calls service management component to carry out the generation of zero-address service, is associated with container group by label, call container group management assembly Generate, pause or delete container group, state controller component carries out all in accordance with determining sequence when executing aforesaid operations, so as into Row control;
In order to realize that the recovery of stateful service, the Reconstruction of The Function of state controller component are existed in stateful service cluster When the associated container group exception of stateful service, state controller component suppressing exception container group first, container group is relevant to hold Longization storage statement, zero-address service will not change, their life cycle and container group is unrelated, therefore use abnormal container group Identical title regenerates container group, the persistent storage statement of the abnormal container group before the container group use regenerated Restore container group state and network identity with zero-address service.
Using aforementioned present invention system and method, it is associated to carry out stateful service in distributed container cluster management platform The relevant operation of container group:
When a stateful service is only associated with a container group, the specific steps of the state controller deployment of components container group:
Step 1:Check whether the container group has carried out the generation and binding of persistent storage statement, if not generating and binding, Step 1 is executed, it is no to then follow the steps 2;
Step 2:It continues checking for whether zero-address service creates, does not create, then execute the zero-address service that creates, it is no to then follow the steps 3;
Step 3:Using fixed container group name, start container group, initialization container group state.
When a stateful N number of container group of service cluster respective associated, state controller component is according to serial number from 0 to N-1 Sequential placement container group when being created, the specific steps are:
Step 1:Whether the container group state for detecting serial number 0 is normal, if normally executing step 2, otherwise deployment container group -0, Then step 1 is executed;
Step 2:Container group serial number adds 1, and whether detection container group state is normal, if normal and serial number is less than N-1, continues to execute Step 2, until container group serial number N-1 then terminates;Otherwise the container group of the serial number is disposed, step 1 is then executed, with reference to figure 4.
The N number of container group of the stateful service cluster respective associated of said one, state controller component according to serial number from N-1 to When 0 sequential placement container group is deleted, the specific steps are:
Step 1:Whether the container group of detection serial number N-1 has deleted, and 2 are thened follow the steps if deleted, and otherwise deletes and holds Device group-N-1 executes step 1;
Step 2:Container group serial number subtracts 1, and whether detection container group has deleted, if deleted and serial number be more than 0, cycle execute Step 2, until container group serial number 0 then terminates;Otherwise the container group of the serial number is deleted, step 1 is executed, with reference to figure 5.
Stateful copy controller assemblies state of a control controller assemblies carry out the operation of stateful service container group capacity reducing When, such as zoom to n container group copy, n from N number of container group copy>0 and n<N confirms the small containers of all serial number ratio n Group must run it is normal with state, the specific steps are:
Step 1:Whether the container group state for detecting serial number 0 is normal, if normally executing step 2, otherwise deployment container group -0, Then step 1 is executed;
Step 2:Container group serial number adds 1, and whether detection container group state is normal, if normal and serial number is less than n, continues to execute step Rapid 2, until container group serial number n-1 thens follow the steps 3;Otherwise the container group of the serial number is disposed, then executes step 1;
Step 3:It checks whether the container group of serial number N-1 has deleted, 4 is thened follow the steps if deleted, otherwise delete and hold Device group-N-1 executes step 1;
Step 4:Container group serial number subtracts 1, and whether detection container group has deleted, if deleted and serial number be more than n-1, execute step Rapid 4, until container group serial number n then terminates;Otherwise the container group of the serial number is deleted, step 3 is executed, with reference to figure 6.
When stateful copy controller assemblies state of a control controller assemblies carry out stateful service container group dilatation operation, Such as extend to N number of container group copy, n from n container group copy>0 and n<N confirms that the small container groups of all serial number ratio n must It must run and state is normal, the specific steps are:
Step 1:Whether the container group state for detecting serial number 0 is normal, if normally executing step 2, otherwise deployment container group -0, Then step 1 is executed;
Step 2:Container group serial number adds 1, and whether detection container group state is normal, if normal and serial number is less than n, continues to execute step Rapid 2, until container group serial number n-1 thens follow the steps 3;Otherwise the container group of the serial number is disposed, then executes step 1;
Step 3:Deployment creates container group, and container group serial number adds 1, starts container group, initialization container group state, detection container group Whether state is normal, if normal and serial number is less than N, continues to execute step 3, until container group serial number N-1 then terminates;Otherwise The container group of the serial number is disposed, then executes step 1, with reference to figure 7.
Illustrate a bit, the starting point of Fig. 4 with Fig. 7 flows is different, there are no the deployment of any container group in Fig. 4, is opened from container group 0 Begin to dispose, Fig. 7 expressions are to have deployed n container group, are disposed since (n+1)th container group, but are needed from container group- 0 starts inspection state.
Using the method for the present invention and system, each container group of stateful service has fixed name, fixed permanentization storage With fixed network address, and the startup of stateful service container group according to name order carry out sequence starting, a container group Start completion and operating status it is normal after restart next container group, it is ensured that the container group of startup can be provided at the distribution of work Source, avoiding starting together causes resource to be seized to cause system unstable and disordering.
And the method for the present invention and system support are run in the container cloud platform of domestic CPU and operating system, Neng Goushi The deployment of standing state service and cluster management, serializing name container group, network identity and persistent storage volume statement, are realized Resource and application logical separation and one-to-one correspondence mutually binding, can carry out stateful service cluster orderly dilatation and orderly contracting Hold, solves the problems, such as the persistence of service state for the stateful service of container cluster.

Claims (10)

1. the realization system of the stateful service of container cluster based on domestic CPU and OS, it is characterized in that including distributed sets of containers The state controller component and stateful copy controller assemblies of group's management platform operation,
Stateful copy controller assemblies are responsible for the stateful service cluster of container cluster, the related stateful service of identification Order executes corresponding stateful service according to the operating condition state of a control controller assemblies of management platform respective resources and is closed Join the operational order of container group, while detecting the state of the container group;
State controller component is responsible for disposing the container group, is named for the container group, provides described in zero-address service conduct The network identity of container group provides persistent storage for the container group and states and bound, and starts to container group suitable Sequence is controlled, while executing specific related behaviour for the container group according to the order of stateful copy controller assemblies Make.
2. system according to claim 1, it is characterized in that the related stateful service of stateful copy controller assemblies identification Order, the container group copy amount associated by a stateful service is excessive, and stateful copy controller assemblies control shape State controller assemblies delete additional container group copy, otherwise stateful copy controller assemblies state of a control controller assemblies are newly-built Container group copy amount, the container group copy amount associated by the stateful service reach preset copy amount.
3. system according to claim 1 or 2, it is characterized in that the stateful copy controller assemblies manage container cluster Stateful service cluster, detect the state of the container group associated by stateful service, the container obtained every the set time The health status of group detects that container group is abnormal, then stateful copy controller assemblies are rebuild different by state controller component The container group that normal container group or the normal container group of dispatch state are replaced, and replaced possesses whole moneys of abnormal container group Source.
4. system according to claim 3, it is characterized in that the appearance associated by the stateful service of state controller deployment of components Device group is renamed according to container group described in the container group name and serial number, and the new name of the container group is used as container Unique identity in cluster, and the order-assigned zero-address service newly named according to the container group, and binding are held Longization stores.
5. system according to claim 4, it is characterized in that state controller component is according to stateful copy controller assemblies Order, call the other assemblies in distributed container cluster management platform to execute life according to determining sequence in certain sequence It enables.
6. system according to claim 5, it is characterized in that state controller component is according to stateful copy controller assemblies Newly-built order, renamed according to container group described in the container group name and serial number for the container group, and according to The sequence that the container group is newly named is named zero-address service and persistent storage, and successively to identical name serial number Container group, zero-address service and persistent storage are bound.
7. system according to claim 5, it is characterized in that state controller component is according to stateful copy controller assemblies Reconstruction order, for the abnormal container group of stateful service, suppressing exception container group, and abnormal container group is rebuild, after reconstruction Container group use the name and zero-address service of abnormal container group and persistent storage.
8. the implementation method of the stateful service of container cluster based on domestic CPU and OS, it is characterized in that utilizing distributed sets of containers The stateful service cluster of the stateful copy controller assemblies management container cluster of group's management platform operation, identification correlation have shape The order of state service executes corresponding stateful according to the operating condition state of a control controller assemblies of management platform respective resources The operational order of the associated container group of service, while detecting the state of the container group;
Container group described in state controller deployment of components using the operation of distributed container cluster management platform is the container group Name, provides network identity of the zero-address service as the container group, and persistent storage statement is provided simultaneously for the container group It is bound, and the sequence started to container group controls, while according to the order of stateful copy controller assemblies, for The container group, executes specific relevant operation.
9. according to the method described in claim 8, it is characterized in that container group described in state controller deployment of components,
When a stateful service is only associated with a container group, specific steps:
Step 1:Check whether the container group has carried out the generation and binding of persistent storage statement, if not generating and binding, Step 1 is executed, it is no to then follow the steps 2;
Step 2:It continues checking for whether zero-address service creates, does not create, then execute the zero-address service that creates, it is no to then follow the steps 3;
Step 3:Using fixed container group name, start container group, initialization container group state;
When a stateful N number of container group of service cluster respective associated, state controller component is according to serial number from 0 to the suitable of N-1 Sequence deployment container group, the specific steps are:
Step 1:Whether the container group state for detecting serial number 0 is normal, if normally executing step 2, otherwise deployment container group -0, Then step 1 is executed;
Step 2:Container group serial number adds 1, and whether detection container group state is normal, if normal and serial number is less than N-1, continues to execute Step 2, until container group serial number N-1 then terminates;Otherwise the container group of the serial number is disposed, step 1 is then executed;Wherein N is Positive integer.
10. according to the method described in claim 8, it is characterized in that the related stateful service of stateful copy controller assemblies identification Order, the container group copy amount associated by a stateful service is excessive, and stateful copy controller assemblies control shape State controller assemblies delete additional container group copy, otherwise stateful copy controller assemblies state of a control controller assemblies are newly-built Container group copy amount, the container group copy amount associated by the stateful service reach preset copy amount.
CN201810411194.2A 2018-05-02 2018-05-02 The implementation method of the stateful service of container cluster based on domestic CPU and OS Pending CN108628613A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810411194.2A CN108628613A (en) 2018-05-02 2018-05-02 The implementation method of the stateful service of container cluster based on domestic CPU and OS

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810411194.2A CN108628613A (en) 2018-05-02 2018-05-02 The implementation method of the stateful service of container cluster based on domestic CPU and OS

Publications (1)

Publication Number Publication Date
CN108628613A true CN108628613A (en) 2018-10-09

Family

ID=63695195

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810411194.2A Pending CN108628613A (en) 2018-05-02 2018-05-02 The implementation method of the stateful service of container cluster based on domestic CPU and OS

Country Status (1)

Country Link
CN (1) CN108628613A (en)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109392020A (en) * 2018-10-26 2019-02-26 中山大学 A kind of resource management platform towards 5G mobile network
CN109491762A (en) * 2018-11-09 2019-03-19 网易(杭州)网络有限公司 Container state control method and device, storage medium, electronic equipment
CN109558260A (en) * 2018-11-20 2019-04-02 北京京东尚科信息技术有限公司 Kubernetes troubleshooting system, method, equipment and medium
CN109710492A (en) * 2018-12-29 2019-05-03 北方工业大学 Application program operation monitoring method, medium and electronic equipment
CN110262944A (en) * 2019-06-21 2019-09-20 四川长虹电器股份有限公司 The method that a kind of pair of K8s cluster container resource is monitored and is alerted
CN110351149A (en) * 2019-07-26 2019-10-18 北京神州绿盟信息安全科技股份有限公司 A kind of method and device for safeguarding network data Forwarding plane
CN111880929A (en) * 2020-07-07 2020-11-03 腾讯科技(深圳)有限公司 Instance management method and device and computer equipment
CN111949379A (en) * 2020-09-18 2020-11-17 成都精灵云科技有限公司 Declarative copy control method based on container scheduling engine
CN112214323A (en) * 2020-10-12 2021-01-12 苏州浪潮智能科技有限公司 Resource recovery method and device and computer readable storage medium
CN113254156A (en) * 2021-05-31 2021-08-13 深信服科技股份有限公司 Container group deployment method and device, electronic equipment and storage medium
CN114490070A (en) * 2022-01-28 2022-05-13 浙江大华技术股份有限公司 Access request forwarding method and device, storage medium and electronic device
CN114584559A (en) * 2022-03-07 2022-06-03 浪潮云信息技术股份公司 Method and device for realizing declarative management of container cloud
US11704165B2 (en) 2021-03-16 2023-07-18 International Business Machines Corporation Persistently available container services through resurrection of user jobs in new compute container instances designated as lead instances

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140101656A1 (en) * 2012-10-10 2014-04-10 Zhongwen Zhu Virtual firewall mobility
CN105245373A (en) * 2015-10-12 2016-01-13 天津市普迅电力信息技术有限公司 Construction and operation method of container cloud platform system
CN105550130A (en) * 2015-12-14 2016-05-04 中电科华云信息技术有限公司 Container based dynamic arrangement method for application environment and system applying method
CN107426034A (en) * 2017-08-18 2017-12-01 国网山东省电力公司信息通信公司 A kind of extensive container scheduling system and method based on cloud platform
CN107844539A (en) * 2017-10-19 2018-03-27 广州阿里巴巴文学信息技术有限公司 Isolated data management method, device, system, computing device and storage medium

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140101656A1 (en) * 2012-10-10 2014-04-10 Zhongwen Zhu Virtual firewall mobility
CN105245373A (en) * 2015-10-12 2016-01-13 天津市普迅电力信息技术有限公司 Construction and operation method of container cloud platform system
CN105550130A (en) * 2015-12-14 2016-05-04 中电科华云信息技术有限公司 Container based dynamic arrangement method for application environment and system applying method
CN107426034A (en) * 2017-08-18 2017-12-01 国网山东省电力公司信息通信公司 A kind of extensive container scheduling system and method based on cloud platform
CN107844539A (en) * 2017-10-19 2018-03-27 广州阿里巴巴文学信息技术有限公司 Isolated data management method, device, system, computing device and storage medium

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
振宇要低调: "Kubernetes有状态应用管理——PetSet", 《HTTPS://WWW.BBSMAX.COM/A/L1DYVVP9ZE/》 *
龚正等: "《Kubernetes权威指南 从Docker到Kubernetes实践全接触》", 30 November 2017, 电子工业出版社 *

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109392020A (en) * 2018-10-26 2019-02-26 中山大学 A kind of resource management platform towards 5G mobile network
CN109491762A (en) * 2018-11-09 2019-03-19 网易(杭州)网络有限公司 Container state control method and device, storage medium, electronic equipment
CN109558260B (en) * 2018-11-20 2022-06-07 北京京东尚科信息技术有限公司 Kubernetes fault elimination system, method, equipment and medium
CN109558260A (en) * 2018-11-20 2019-04-02 北京京东尚科信息技术有限公司 Kubernetes troubleshooting system, method, equipment and medium
CN109710492A (en) * 2018-12-29 2019-05-03 北方工业大学 Application program operation monitoring method, medium and electronic equipment
CN110262944A (en) * 2019-06-21 2019-09-20 四川长虹电器股份有限公司 The method that a kind of pair of K8s cluster container resource is monitored and is alerted
CN110351149A (en) * 2019-07-26 2019-10-18 北京神州绿盟信息安全科技股份有限公司 A kind of method and device for safeguarding network data Forwarding plane
CN111880929A (en) * 2020-07-07 2020-11-03 腾讯科技(深圳)有限公司 Instance management method and device and computer equipment
CN111880929B (en) * 2020-07-07 2024-02-02 腾讯科技(深圳)有限公司 Instance management method and device and computer equipment
CN111949379A (en) * 2020-09-18 2020-11-17 成都精灵云科技有限公司 Declarative copy control method based on container scheduling engine
CN112214323A (en) * 2020-10-12 2021-01-12 苏州浪潮智能科技有限公司 Resource recovery method and device and computer readable storage medium
CN112214323B (en) * 2020-10-12 2022-06-14 苏州浪潮智能科技有限公司 Resource recovery method and device and computer readable storage medium
US11704165B2 (en) 2021-03-16 2023-07-18 International Business Machines Corporation Persistently available container services through resurrection of user jobs in new compute container instances designated as lead instances
CN113254156A (en) * 2021-05-31 2021-08-13 深信服科技股份有限公司 Container group deployment method and device, electronic equipment and storage medium
CN113254156B (en) * 2021-05-31 2024-04-09 深信服科技股份有限公司 Container group deployment method and device, electronic equipment and storage medium
CN114490070A (en) * 2022-01-28 2022-05-13 浙江大华技术股份有限公司 Access request forwarding method and device, storage medium and electronic device
CN114584559A (en) * 2022-03-07 2022-06-03 浪潮云信息技术股份公司 Method and device for realizing declarative management of container cloud
CN114584559B (en) * 2022-03-07 2024-07-12 浪潮云信息技术股份公司 Device for realizing declaration type management of container cloud

Similar Documents

Publication Publication Date Title
CN108628613A (en) The implementation method of the stateful service of container cluster based on domestic CPU and OS
CN112118565B (en) Multi-tenant service gray level publishing method, device, computer equipment and storage medium
CN112667362B (en) Method and system for deploying Kubernetes virtual machine cluster on Kubernetes
JP4762972B2 (en) System and method for a recovery unit in a database
Nathan et al. Comicon: A co-operative management system for docker container images
US8433948B2 (en) Method and apparatus for realizing application high availability
CN112099918A (en) Live migration of clusters in containerized environments
US8527558B2 (en) Distributed garbage collection
CN109391664A (en) System and method for the deployment of more cluster containers
CN1329840C (en) File archival
CN103827822A (en) System and method for supporting automatically deploying/undeploying application components in a transactional middleware machine environment
CN102355499B (en) Cloud computing system
CN111324538B (en) Dynamic routing-based micro-service parallel test environment management method
CN109684028B (en) Method, device and equipment for separating operating system from user data
CN109165206B (en) High-availability implementation method for HDFS (Hadoop distributed File System) based on container
CN115129736A (en) Rule engine-based rule event dynamic loading and updating method and related equipment
CN113656149B (en) Application processing method and device and related equipment
US12026066B2 (en) Backing up data for a namespace assigned to a tenant
CN113918281A (en) Method for improving cloud resource expansion efficiency of container
Oliveira et al. Delivering software with agility and quality in a cloud environment
CN110134489A (en) Using moving method and device, system, storage medium, using upper cloud tool
CN107180034A (en) The group system of MySQL database
CN110494849A (en) For determining the successful system and method for cross-platform migration
Trunov et al. Legacy applications model integration to support scientific experiment
CN110958138B (en) Container expansion method and device

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20181009

RJ01 Rejection of invention patent application after publication