KR102567541B1 - Edge service instance distribution apparatus and control method thereof - Google Patents

Abstract

The present invention relates to an edge service instance distribution device and a control method thereof. A control method of an edge service instance distribution apparatus according to the present invention includes distributing a new version edge service instance corresponding to a predetermined service; checking an operating state of the new version edge service instance; As a result of the check, when a failure occurs in the operation of the corresponding new version edge service instance, redistributing the new version edge service instance where the corresponding failure occurs; After the redistribution, checking the operating state of the redistributed new version edge service instance; determining the number of redistributions of the new version edge service instance in which the failure occurs, when a failure occurs in the operation of the corresponding redistributed new version edge service instance as a result of the check; As a result of the determination, when the number of redistribution is equal to or greater than the preset redistribution reference value, the method may include a step of distributing an edge service instance of a previous version and controlling rollback processing.

Description

Edge service instance distribution device and its control method {EDGE SERVICE INSTANCE DISTRIBUTION APPARATUS AND CONTROL METHOD THEREOF}

The present invention relates to an edge service instance distribution apparatus and a control method thereof, and more particularly, to an apparatus and a control method for distributing a new edge service instance as an edge service is upgraded (updated).

Cloud computing, also known as cloud computing, refers to the provision of computing services such as servers, storage, software, and analytics over the "cloud," or the Internet. For example, Naver's Endlive, Google Docs, and OneDrive, which are installed together with Windows, are representative examples of such services, and many users are currently using cloud functions easily.

However, cloud computing also has several problems.

As the number of people using cloud services increases exponentially, they begin to exceed the amount of data that can be processed by servers and data centers, and in addition, data delays occur in the process of analyzing and transmitting the collected data. .

Edge computing has been introduced as one of methods to solve problems related to the conventional cloud computing, particularly data processing speed and capacity.

Edge computing, also called fog computing or MEC (Mobile Edge Computing), refers to computing at the end device, that is, the 'edge'.

This means that the computing device is located close to the terminal devices used by users, rather than a place where data is processed and computed in a far away data center like cloud computing.

In particular, if equipment located at the 'edge' so far only plays the role of data transmission or data storage, edge computing adds computing power to this 'edge' equipment to provide data analysis or other various services. there is.

Such edge computing can be used as a modified form of cloud rather than being configured alone, and this can be referred to as cloud edge computing.

However, various services may be provided in a system created to implement such cloud edge computing, that is, a cloud edge system. For example, a service provision base may be a 'container'.

Here, a container is a kind of virtualization technology, and since it includes data, programs, and information necessary for providing each service, the same container is widely used because it can perform the same function regardless of the version of the server operating system.

Each of the services provided in the cloud edge system may be referred to as an edge service or an edge service instance, and each provides a service while consuming resources in the cloud edge system.

Edge service instances corresponding to each of these services may be periodically or intermittently replaced with new versions.

For example, if a new version of the edge service instance is registered while an edge service instance corresponding to a message providing service is installed, replace or renew the existing (old version) edge service instance with the new version edge service instance. It can be.

This may correspond to version management of general applications.

However, after a new version is created and the edge service instance is deployed with the new version, if a failure occurs while the edge service instance of the new version is in operation, the new version of the edge service instance can be redeployed to install and operate the instance. can

This means that due to the operation of the edge service instance for a long time, some functions may fail in the internal processing process, and accordingly, the overall functional operation of the edge service instance may not be performed. Finding or searching is a very cumbersome and time-consuming process, so in the past, a new version of the edge service instance was easily redistributed to install and operate.

However, if the function of the new version of the edge service instance repeatedly fails, another problem arises as the edge service instance of the same version (ie, the new version) is infinitely redeployed.

That is, as described above, if the failure that occurs during the operation of the updated edge service instance is caused by a problem inherent in the edge service instance itself (package itself) of the corresponding version, even if the edge service instance of the same version is redistributed. The same failure may occur repeatedly, resulting in harming the continuity of service operation.

Therefore, it is requested to propose a method to solve this problem related to the redistribution of the updated edge service instance.

Publication No. 10-2021-0095457

The present invention has been devised to meet the above-mentioned conventional request, and its purpose is to redeploy an edge service instance, a device for preventing repetition of failure due to a problem inherent in the edge service instance itself and its control is to provide a way

In order to achieve the above object, a control method of an edge service instance distribution device according to the present invention includes distributing a new version edge service instance corresponding to a predetermined service; checking an operating state of the new version edge service instance; As a result of the check, when a failure occurs in the operation of the corresponding new version edge service instance, redistributing the new version edge service instance where the corresponding failure occurs; After the redistribution, checking the operating state of the redistributed new version edge service instance; determining the number of redistributions of the new version edge service instance in which the failure occurs, when a failure occurs in the operation of the corresponding redistributed new version edge service instance as a result of the check; As a result of the determination, when the number of redistribution is equal to or greater than a preset redistribution reference value, a step of distributing an edge service instance of a previous version and controlling rollback processing may be included.

Here, as a result of the determination, if the number of redistribution times is less than a preset redistribution reference value, the corresponding new version edge service instance may be redistributed.

Here, the initial distribution time of the new version edge service instance may be stored, and the redistribution reference value may be dynamically changed in proportion to the elapsed time from the initial distribution time for each of the plurality of new version edge service instances.

Here, after performing periodic access to the socket communication corresponding to the new version edge service instance, it is possible to determine whether or not a failure has occurred according to the result.

In order to achieve the above object, an apparatus for distributing an edge service instance according to the present invention includes an operating state checking unit for checking the operating state of an edge service instance installed on its own or remotely; and a new version edge service instance corresponding to a predetermined service. After deploying, if it is confirmed that a failure has occurred as a result of checking the operating state of the new version edge service instance in the operation status check unit, redistribution of the new version edge service instance with the corresponding failure occurs, and the redistributed new version If it is confirmed by the operation status check unit that the operation state of the edge service instance has failed, the number of redistributions of the new version edge service instance where the failure has occurred is determined, and as a result of the determination, if the number of redistributions is greater than the preset redistribution reference value, the previous version edge It can be configured to include an instance distribution control unit that deploys service instances and performs rollback processing control.

Here, the instance distribution control unit may redistribute the corresponding new version edge service instance when the number of redistribution is less than a preset redistribution reference value.

Here, a distribution time management unit for storing the initial distribution time of the new version edge service instance by the instance distribution control unit; and dynamically changing the redistribution reference value in proportion to the elapsed time from the initial distribution time for each of a plurality of new version edge service instances. It may further include a setting value changer to do.

Here, the operating status check unit may perform periodic access to socket communication corresponding to each edge service instance and determine whether or not a failure has occurred according to the result.

As described above, according to the present invention, if a failure of the corresponding edge service instance repeatedly occurs after a new version of the edge service instance is deployed, or if the preset redistribution threshold value is exceeded, the edge service instance of the previous version is deployed again. By making rollback by doing so, problems caused by errors in the package itself can be minimized.

In particular, by dynamically changing the redistribution reference value according to the elapsed time from the initial distribution, even if it is not specifically determined whether it is a package problem or another hardware problem, edge service instances that do not have a package problem with a high probability are continuously redistributed, and Anything that has problems can be rolled back to a previous version.

1 is a schematic configuration diagram of an entire system including an edge service instance distribution device according to an embodiment of the present invention;
2 is a functional block diagram of the edge service instance distribution device of FIG. 1;
3 is an overall control flow diagram of an edge service instance distribution device according to an embodiment of the present invention.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

Hereinafter, each embodiment according to the present invention is only one example to aid understanding of the present invention, and the present invention is not limited to these embodiments. In particular, the present invention may be composed of at least one or more combinations of individual components, individual functions, or individual steps included in each embodiment.

In particular, although alphabets such as '(a)' are included in some claims of the claims for convenience, these alphabets do not prescribe the order of each step.

A schematic configuration of the entire system including the edge service instance distribution device 100 according to an embodiment of the present invention is as shown in FIG.

As shown in the figure, the entire communication system may include an edge service instance distribution device 100 and an edge server 200.

First of all, the edge server 200 includes at least one edge service instance to provide a specific service to a user or customer. Here, the edge service instance refers to each module for performing various services provided in the cloud edge system. However, since this corresponds to a known technique, a detailed description thereof will be omitted.

A plurality of such edge service instances may be driven in one edge server 200 .

Meanwhile, the edge service instance distribution device 100 performs a function of distributing edge service instances required for the edge server 200 while communicating with each edge server 200 .

In this embodiment, the edge server 200 and the edge service instance distribution device 100 are physically separated as an example, but in fact, the functions of the edge service instance distribution device 100 and the edge server ( 200) may be included as a matter of course.

2, specific functional blocks of the edge service instance distribution device 100 are disclosed.

As shown in the figure, the edge service instance distribution device 100 is configured to include an operating status check unit 110, an instance distribution control unit 120, a distribution time management unit 130, and a setting value change unit 140. can

The operating state checking unit 110 performs a function of checking the operating state of an edge service instance installed on its own or remotely installed.

For example, when an edge service instance is installed in the edge service instance distribution device 100, the operating state check unit 110 checks the operating state of the edge service instance itself installed, and as another example, the remote edge server 200 When an edge service instance is installed in , the operating state of the edge service instance of the corresponding edge server 200 is checked.

That is, the operating state check unit 110 performs periodic access to socket communication corresponding to each edge service instance, and then determines whether or not a failure has occurred according to the result.

As a method of checking the operating state of the edge service instance, a method of transmitting a dummy packet to a port used by the corresponding edge service instance at a preset time interval and then determining whether a normal response is received may be used. , At this time, it can be accessed on a web basis or directly communicated through socket communication.

Furthermore, the operating state of the corresponding edge service instance may be checked by determining the ratio between the request to the corresponding edge service instance and the response coming from it.

The instance distribution control unit 120 performs a function of distributing edge service instances of various versions corresponding to a predetermined service.

For example, if a first version (old version) and a second version (updated version or new version) exist for a specific edge service instance, either one of them can be deployed according to circumstances.

Since the distribution of edge service instances per se corresponds to a known technology, a detailed description thereof will be omitted.

Hereinafter, the process of determining which version of edge service instance to deploy for each situation will be described in detail.

If a new version exists for a given edge service instance, the instance distribution control unit 120 may preferentially distribute the new version edge service instance.

That is, the instance distribution control unit 120 can deploy the new version edge service instance to be installed and run for the edge server 200 that has never performed the corresponding edge service instance, and already has the previous version of the corresponding edge service instance. The edge server 200 that is running can be controlled to be updated (updated) with a new version of the edge service instance.

After deploying the new version edge service instance, if it is confirmed that a failure has occurred as a result of checking the operating state of the new version edge service instance in the operation status check unit 110, the instance distribution control unit 120 checks the new version edge service instance where the failure has occurred. Perform redeployment of version edge service instances.

Here, failure in the operating state of the new version edge service instance does not mean only a bug in the corresponding edge service instance itself, but includes all cases where the operating state of the new version edge service instance is not smooth for any reason. do.

For example, it includes all cases where some data of a new version edge service instance that was deployed/redistributed due to hardware load is damaged.

This redistribution process may be repeated.

That is, whenever a failure occurs in the operating state of the redistributed new version edge service instance, the instance distribution controller 120 repeatedly redistributes the corresponding new version edge service instance.

However, there is a limitation in this case. The instance distribution controller 120 may determine the number of redistributions and perform other processing according to the determined number of redistributions.

That is, the instance distribution control unit 120 determines the number of redistributions of new version edge service instances that have failed, and redistributes new version edge service instances again if the number of redistributions is less than the preset redistribution reference value.

However, as a result of the determination, if the number of redistribution is equal to or greater than the predetermined reference value, the instance distribution control unit 120 distributes the edge service instance of the previous version (the above-described first version) and controls the rollback processing.

Accordingly, by repeatedly redistributing edge service instances of a new version just because the version is higher, it is possible to minimize a situation in which an actual service is not performed.

In other words, the failure of the new version edge service instance may be due to an error in the package itself, but in the case of repeatedly redistributing only the new version, the service interruption time may rather increase. The present invention minimizes the occurrence of such problems It can be done.

Meanwhile, the distribution time management unit 130 performs a function of storing the initial distribution time of the new version edge service instance by the instance distribution control unit 120 .

For example, when a new version edge service instance is distributed to a plurality of edge servers 200, the distribution time management unit 130 stores the first distribution time among them.

The setting value changing unit 140 performs a function of dynamically changing the redistribution reference value in proportion to the elapsed time from the initial distribution time for each of the plurality of new version edge service instances.

As described above, the instance distribution control unit 120 compares the number of redistributions with a preset redistribution reference value when determining whether to redistribute when an operation failure of a new version edge service instance occurs. is to dynamically change the redistribution reference value in proportion to the elapsed time from the initial distribution of the corresponding new version edge service instance.

As a result, the redistribution threshold is relatively increased for new version edge service instances (eg, new version edge service instances for search services) that have been first deployed for a long time, and the new version not long after the initial deployment. For an edge service instance (eg, a new version edge service instance for a messenger service), the redistribution reference value becomes relatively small.

This means that edge service instances where a long time has elapsed from the time of initial distribution have been fully analyzed for the cause of the error, and therefore, it is highly likely that the error in the package itself is not the case, so additional redistribution is performed.

However, it is highly likely that edge service instances that have not yet been deployed have properly analyzed packet errors, so it is preferable to roll back to the previous version after several redeployments in terms of service maintenance.

Hereinafter, a control flow of the edge service instance distribution device 100 according to an embodiment of the present invention will be described with reference to FIG. 3 .

First, the edge service instance distribution device 100 registers a management policy for each edge service instance (step S1). Since this policy registration itself corresponds to a known technology, a detailed description thereof will be omitted.

Here, the management policy may include information for detecting a failure state (port number for health check, etc.), and the number of allowed redistribution as information for a rollback criterion.

The edge service instance distribution device 100, when registration of a new version is detected for a specific edge service instance (step S3), deploys the corresponding new version edge service instance (step S5).

Thereafter, the edge service instance distribution device 100 continuously monitors the operating state of the corresponding new version edge service instance and checks the service failure state (step S7).

When a service failure state is confirmed, the edge service instance deployment device 100 compares the number of allowed redistribution of the corresponding edge service instance with the number of actual redistribution (step S), and if the result does not exceed the allowed number of redistribution, redistribution is performed again. The new version edge service instance is redeployed (step S11).

If, as a result of the comparison, the actual number of redistribution exceeds the allowable number of redistribution (step S9), the edge service instance distribution device 100 distributes the edge service instance of the previous version to control rollback processing (step S13).

Accordingly, if there is a problem with the new version edge service instance package itself, it is quickly rolled back to the edge service instance of the previous version, enabling smooth service operation.

Meanwhile, it goes without saying that the process of performing each of the above-described embodiments may be performed by a program or application stored in a predetermined recording medium (for example, computer-readable). Here, the recording medium includes all of an electronic recording medium such as RAM (Random Access Memory), a magnetic recording medium such as a hard disk, an optical recording medium such as a CD (Compact Disk), and the like.

At this time, the program stored in the recording medium may be executed on hardware such as a computer or smart phone to perform each of the above-described embodiments. In particular, at least one of the functional blocks of the edge service instance distribution device 100 according to the present invention described above may be implemented by such a program or application.

In addition, the present invention is not limited to the specific embodiment described above, but can be implemented by various modifications and variations within the scope of the present invention. It will be apparent that such variations and modifications are included in the present invention provided they come within the scope of the appended claims.

100: Edge service instance distribution device 200: Edge server
110: operation status check unit 120: instance distribution control unit
130: distribution time management unit 140: set value change unit

Claims (10)

(a) distributing a new version edge service instance corresponding to a predetermined service;
(b) checking an operating state of the new version edge service instance;
(c) when a failure occurs in the operation of the corresponding new version edge service instance as a result of checking in step (b), redistributing the new version edge service instance where the corresponding failure occurred;
(d) after the redistribution in step (c), checking the operation state of the redistributed new version edge service instance;
(e) determining the number of redistributions of the new version edge service instance where the failure occurred, if a failure occurred in the operation of the redeployed new version edge service instance as a result of checking in step (d);
(f) if the number of redistributions as a result of the determination in step (e) is equal to or greater than the preset redistribution reference value, distributing the previous version edge service instance to control rollback processing;
In step (a), the initial distribution time of the new version edge service instance is stored,
The redistribution reference value is dynamically changed in proportion to the elapsed time from the initial distribution for each of the plurality of new version edge service instances, but the redistribution reference value for the new version edge service instances for which a longer time than the preset time has elapsed since the initial distribution. The redistribution reference value increases according to a preset criterion, and the redistribution reference value decreases according to a preset criterion for a new version edge service instance within a preset time after being first deployed. control method. According to claim 1,
In the step (f), if the number of redistributions is less than a predetermined redistribution reference value as a result of the determination in the step (e), the corresponding new version edge service instance is redistributed. delete According to claim 1,
In the step (b), after performing periodic access to the socket communication corresponding to the new version edge service instance, it is characterized in that it is determined whether a failure occurs according to the result. A computer-readable recording medium recording a program for executing the method of any one of claims 1, 2, and 4. An application program stored in a computer readable recording medium in order to execute the method of any one of claims 1, 2, and 4 in combination with hardware. an operating state checking unit that checks the operating state of an edge service instance installed on its own or remotely;
After deploying a new version edge service instance corresponding to a predetermined service, if it is confirmed that a failure has occurred as a result of checking the operating state of the new version edge service instance in the operation status check unit, the new version edge service instance with the corresponding failure has occurred. redistribution is performed, and if it is confirmed by the operation status checking unit that the operation state of the redistributed new version edge service instance has a failure, the number of redistributions of the new version edge service instance with the failure is determined, and as a result of the determination, the number of redistributions Including an instance distribution control unit that deploys an edge service instance of a previous version when the value is greater than the preset redistribution reference value and controls rollback processing;
a distribution time management unit for storing the initial distribution time of the new version edge service instance by the instance distribution control unit;
For each of the plurality of new version edge service instances, the redistribution reference value is dynamically changed in proportion to the elapsed time from the initial distribution. It is characterized by further comprising a setting value changing unit that increases the redistribution reference value according to a preset standard and reduces the redistribution reference value according to a preset standard for a new version edge service instance within a preset time after the first deployment. edge service instance deployment device. According to claim 7,
Wherein the instance distribution control unit redistributes the corresponding new version edge service instance when the number of redistributions is less than a preset redistribution reference value. delete According to claim 7,
The operation state check unit performs periodic access to socket communication corresponding to each edge service instance, and then determines whether a failure occurs according to the result.