CN110311831B

CN110311831B - Container cloud-based system resource monitoring method and related equipment

Info

Publication number: CN110311831B
Application number: CN201910515745.4A
Authority: CN
Inventors: 高峰
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2019-06-14
Filing date: 2019-06-14
Publication date: 2022-03-25
Anticipated expiration: 2039-06-14
Also published as: WO2020248507A1; CN110311831A

Abstract

The invention relates to the technical field of system resource monitoring, in particular to a system resource monitoring method based on a container cloud and related equipment, wherein the method comprises the following steps: generating a frame list after acquiring the deployment condition of the container layout frame; acquiring running state information of the application in each container arrangement frame; determining that the container arrangement frame resources are insufficient according to the running state information, generating alarm information, and pushing the alarm information to a capacity expansion executor; acquiring physical machine resource configuration data marked as a resource-deficient container arrangement frame and physical machine resource occupation data occupied by application, and then recording the data in a corresponding recording node; after the capacity expansion is finished, the application is reconfigured and restarted, and the current physical machine resource configuration data of the container arrangement framework is obtained; and generating a capacity expansion report. According to the invention, the running state of the application on the container cloud platform is monitored, early warning is timely carried out when the system resources are insufficient, the capacity expansion requirement of the container arrangement framework is quickly responded, and historical data before and after capacity expansion is reserved.

Description

Container cloud-based system resource monitoring method and related equipment

Technical Field

The invention relates to the technical field of system resource monitoring, in particular to a method, a device, equipment and a storage medium for monitoring system resources based on a container cloud.

Background

With the gradual popularization of the application of distributed computing resources, a container cloud technology is favored by various users, and an internet cloud computing service provider also develops own products through the container cloud technology according to own characteristics, for example, an airy cloud and an Tencent cloud integrating the container cloud technology in a large product series of own homes, or a safety platform developed in a deep customization manner, and the like, wherein the products are distributed platforms based on an application container engine Docker, and can complete the rapid creation, operation, rapid capacity reduction and expansion and fault self-healing of an application program. The use of these container cloud platforms requires the allocation and management of resources for various types of services and applications running on the platform by relying on the container orchestration framework. Such as Docker Swarm, Marathon, kubernets, Nomad, etc. based on Docker. The resources of the individual services and applications are allocated in a rational manner by means of these orchestration tools and are recovered when an application or service crashes. Common container layout framework products provide a friendly interface, a RestAPI and other simple and easy-to-use data interfaces to create and manage applications, and have the convenience of integration with a third-party system, for example, a Marathon framework can also define applications or services through a JSON format text, and the application definition is submitted and run through the RestAPI after the application definition is completed, so that the use difficulty of the application is greatly reduced.

In the conventional scheme in the industry, along with the duration of the service time and the expansion of the service, the capacity of system resources is often expanded on the same platform under the condition that the original deployment structure is not greatly adjusted, so as to meet the continuously increasing demand of the application or service corresponding to the developing service on the system resources. For example, a plurality of Marathon frameworks are correspondingly built and deployed on a Padis platform according to different service types to form a Marathon cluster, and the framework sets manage various applications or services running in different service types. With the continuous development of services, the situation that the system resources occupied by the existing applications are in shortage, so that the applications run slowly and even crash is often caused, at this time, even if the applications are restarted, the completion is not done, and at this time, the capacity expansion of the system resources needs to be performed on the Marathon framework where the applications are located in time. However, in the prior art, the container monitoring tool cAdvisor of Google is generally used to view the use situation of physical machine resources occupied by each application or service running on a container layout framework such as Marathon, and such technical means have the following limitations:

1) only one physical host can be monitored at the same time, which is equivalent to single-node monitoring and cannot meet the requirement of multi-node monitoring, however, applications running on the same container cloud platform may be distributed to run in machine resources managed by different container arrangement frameworks, and therefore may run on different physical hosts, and single-node monitoring cannot meet the monitoring requirement of actual resource usage of the applications.

2) The method can only check the state in real time, and cannot check historical data, so that historical data support cannot be provided for some functions for analyzing the operation trend of the application and the service on the container cloud platform.

3) The early warning function is weak, and the function of telephone or mail warning is lacked, so that the container arrangement framework cannot give an early warning to the outside in time when the physical machine resources are insufficient, and in the actual operation process of the container cloud platform, especially when a certain application is restarted or created, the insufficient physical machine resources can cause that the application cannot be started or the creation is successful, and if the application cannot be processed in time, the business function corresponding to the application is paralyzed.

Therefore, a technical means for conveniently performing multi-node monitoring, historical data viewing and analysis and fault early warning on the used resources of the container arrangement framework in the container cloud platform is needed in the industry to solve the technical problems.

Disclosure of Invention

The invention provides a system resource monitoring method and a related device based on a container cloud, which are characterized in that monitoring nodes are arranged for each container arrangement frame in a container cloud platform to acquire real-time state information of application, whether the container arrangement frame has insufficient resources is determined by judging the state information, capacity expansion early warning is sent out, and timely application restart is carried out after capacity expansion according to backup of resource configuration state of the application before capacity expansion, so that the effects of multiple monitoring nodes, historical data retention and automatic early warning which cannot be achieved by a monitoring means in a traditional container cloud platform are achieved.

In a first aspect, the present invention provides a system resource monitoring method based on a container cloud, including: acquiring the deployment condition of a container arrangement framework under a container cloud platform, and generating a framework list, wherein all container arrangement frameworks deployed under the container cloud platform are recorded in the framework list;

acquiring the running state information of each application in each container arrangement frame one by one from the frame list according to a recording sequence and a preset acquisition period, and recording the acquired running state information in a preset storage unit, wherein a frame recording node for recording physical machine resource configuration data of each container arrangement frame and an application recording node for recording the running state information of each application and physical machine resource occupation data are arranged in the storage unit, and the running state information is used for identifying the running state of the application in the container arrangement frame in which the running state information is located;

when the running state information of any application is continuously in a waiting state within a preset judgment time threshold range, marking that the resources of the container arrangement frame are insufficient, and at the moment, generating alarm information and pushing the alarm information to an executor executing capacity expansion operation so as to conveniently inform the executor to execute capacity expansion operation of the container arrangement frame;

acquiring physical machine resource configuration data marked as a resource-deficient container arrangement frame and physical machine resource occupation data occupied by any application running in the container arrangement frame, and recording the two types of data in corresponding recording nodes;

after receiving an expansion operation ending signal fed back by the executor, calling previously recorded physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency from the storage unit, then reconfiguring and restarting the application, and recording the current physical machine resource configuration data of the container arrangement frame marked as resource deficiency in the frame recording node after obtaining the current physical machine resource configuration data;

and summarizing the recorded data of the frame recording nodes and the application recording nodes and then generating a capacity expansion report.

In some possible embodiments, the generating a frame list after obtaining the deployment condition of the container orchestration frame under the container cloud platform, where the frame list records all container orchestration frames deployed under the container cloud platform, includes:

a management console connected to the container cloud platform;

sending a data request for acquiring the condition of a container orchestration framework running on the container cloud platform to a management console of the container cloud platform;

receiving feedback of the management console and then generating the frame list, wherein all container arrangement frames running on the container cloud platform are recorded in the frame list according to the feedback time sequence;

and generating a recording sequence number for each container arrangement frame in the frame list according to the recording time, wherein the recording sequence number is an identification sequence number of the container arrangement frame in the container cloud platform and is used for distinguishing different container arrangement frames.

In some possible embodiments, the acquiring the running state information of each application in each container arrangement frame one by one from the frame list according to the recording order according to a preset acquisition period, and recording the acquired running state information in a preset storage unit includes:

generating a monitoring node for each container arrangement frame in the frame list, wherein the monitoring node is used for acquiring running state information of each application running on the monitoring node after being connected with a management console of the container arrangement frame in a set period;

generating a corresponding application recording node for the application on each container arrangement frame in the storage unit according to the recording sequence number of the container arrangement frame in the frame list, wherein the application recording node is used for recording the running state information of each application running on the container arrangement frame, which is acquired by the monitoring node;

through the monitoring node, after the management console of the container arrangement frame is connected according to a set monitoring period, requesting to acquire running state information of all applications running on the container arrangement frame;

and after receiving the feedback of the management console of the container arrangement framework, recording the running state information of the application in the application recording node according to the time when the feedback is received.

In some possible embodiments, when the running state information of any one of the applications continues to be in a waiting state within a preset determination time threshold, it is marked that the container arrangement frame is short of resources, and at this time, after generating the alarm information and pushing the alarm information to an executor executing the capacity expansion operation, the executor is convenient to notify the executor of executing the capacity expansion operation of the container arrangement frame, where the method includes:

reading running state information of any application in the application recording nodes;

judging whether the running state information applied in the judging time threshold range is in a waiting state continuously, if so, marking the state of the container arrangement frame as insufficient resources, and if not, marking the state of the container arrangement frame as normal running, wherein the judging time threshold range is a preset period of time;

traversing all applications under all container arrangement frames in the frame list according to the steps, and marking the states of all container arrangement frames;

generating an alarm mail after calling a mail template, and recording a recording serial number of the container arrangement frame marked as resource deficiency and prompt information for marking resource deficiency in the alarm mail;

and after reading the mail address of the executor from a preset addressee address list, pushing the alarm mail to the executor.

In some possible embodiments, the obtaining physical machine resource configuration data of the container arrangement frame marked as resource deficiency and physical machine resource occupation data occupied by any application running in the container arrangement frame records the two types of data in corresponding recording nodes, including:

a management console connecting the container orchestration framework marked as resource-deficient;

sending a data request to the management console, wherein the data request is used for acquiring physical machine resource configuration data of the container arrangement frame and physical machine resource occupation data occupied by any application running in the container arrangement frame;

and after receiving the feedback of the management console, recording the physical machine resource configuration data in the frame recording node and recording the physical machine resource occupation data to the application recording node according to the time of receiving the feedback.

In some possible embodiments, after receiving the expansion job end signal fed back by the executor, the method retrieves, from the storage unit, previously recorded physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency, performs reconfiguration and restart of the application, acquires current physical machine resource configuration data of the container arrangement frame marked as resource deficiency, and records the current physical machine resource configuration data in the frame recording node, including:

receiving feedback information of the executor, wherein the feedback information comprises a capacity expansion operation ending signal;

after the storage unit is connected, reading a record from the application recording node, wherein the record is the closest to the current time of the physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency;

configuring a corresponding application according to the physical machine resource occupation data, and restarting the application after configuration is completed;

and the management console connected with the container arrangement frame acquires the current physical machine resource configuration data of the container arrangement frame and records the acquired data in the frame recording node according to the acquisition time.

In some possible embodiments, after generating the frame list after acquiring the container orchestration frame deployment situation under the container cloud platform, the method includes:

connecting the management control console of each container arrangement frame one by one according to the recording sequence in the frame list;

arranging a frame for the containers which are successfully connected, and generating a new recording sequence number after adding a successful mark after the recording sequence number;

arranging a frame for the container with failed connection, and generating a new recording sequence number after adding a failure mark behind the recording sequence number;

the method comprises the steps of acquiring running state information of each application in each container arrangement frame one by one from the frame list according to a recording sequence according to a preset acquisition period, identifying a recording sequence number of the container arrangement frame before recording the acquired running state information in a preset storage unit, reading the running state information of the application in the container arrangement frame after the recording sequence number contains a success mark, and not executing the operation of reading the running state information of the application in the container arrangement frame when the recording sequence number contains a failure mark.

In a second aspect, the present invention provides, in some possible embodiments, a container cloud-based system resource monitoring apparatus, including: the device comprises a list generation module, an application state acquisition module, an alarm information pushing module, a data recording module, an application restarting module and an expansion report generation module, wherein:

the list generation module is used for generating a frame list after acquiring the container arrangement frame deployment condition under the container cloud platform;

the application state acquisition module is arranged for acquiring the running state information of each application in each container arrangement frame one by one from the frame list according to a recording sequence and a preset acquisition period, and recording the acquired running state information in a preset storage unit;

the alarm information pushing module is set to mark that the container arrangement frame resources are insufficient when the running state information of any one application is continuously in a waiting state within a preset judgment time threshold range, and pushes the running state information to an executor executing capacity expansion operation after alarm information is generated;

the data recording module is set to acquire physical machine resource configuration data marked as a container arrangement frame with insufficient resources and physical machine resource occupation data occupied by any application running in the container arrangement frame, and records the two types of data in corresponding recording nodes;

the application restarting module is configured to receive a capacity expansion operation ending signal fed back by the executor, retrieve previously recorded physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency from the storage unit, perform application reconfiguration and restart, acquire current physical machine resource configuration data of the container arrangement frame marked as resource deficiency, and record the current physical machine resource configuration data in the frame recording node;

and the capacity expansion report generation module is configured to generate a capacity expansion report after summarizing the recorded data of the frame recording nodes and the application recording nodes.

Based on the same inventive concept, the present invention provides, in some possible embodiments, a computer device including a memory and a processor, wherein the memory stores computer readable instructions, and the computer readable instructions, when executed by the processor, implement the steps of the container cloud based system resource monitoring method described above.

Based on the same inventive concept, the present invention provides, in some possible embodiments, a computer-readable storage medium having stored thereon computer-readable instructions which, when executed by one or more processors, implement the steps of the container cloud based system resource monitoring method described above.

Has the advantages that: the invention monitors the running states of all applications running on the container cloud platform, and gives an early warning in time when the system resources are insufficient, so that the capacity expansion requirement of the container arrangement framework where the applications are located is quickly and accurately responded, and historical data before and after capacity expansion is reserved for subsequent analysis and query, and the invention has the following advantages:

1) multi-node monitoring: starting from a container arrangement framework on a container cloud platform, monitoring various applications by calling console data, and judging the system resource condition by acquiring the running state of the applications, wherein the monitoring process does not depend on physical hardware;

2) automatic early warning: by a built-in automatic early warning mechanism, under the condition that the system resource is judged to be insufficient, early warning information is sent out in time, and the efficiency of capacity expansion countermeasures is improved;

3) the historical data can be looked up: the physical machine resource configuration data of the container arrangement framework before and after capacity expansion, the application state information, the occupation condition of the physical machine resources and other data are recorded to respective recording nodes, so that the effect of inquiring historical data is realized, and data support is provided for the system resource change trend caused by the pre-configuration and the business development of the application on the used resources conveniently and subsequently.

Drawings

Fig. 1 is a main flow chart of a container cloud-based system resource monitoring method according to an embodiment of the present invention;

fig. 2 is a flowchart of generating a frame list in a container cloud-based system resource monitoring method according to an embodiment of the present invention;

fig. 3 is a flowchart of monitoring an application state in a container cloud-based system resource monitoring method according to an embodiment of the present invention;

fig. 4 is a flowchart illustrating a resource shortage determination method in a system resource monitoring method based on a container cloud according to an embodiment of the present invention;

fig. 5 is a flowchart of data backup before capacity expansion in a system resource monitoring method based on a container cloud according to an embodiment of the present invention;

fig. 6 is a flowchart illustrating recovery of application operation after capacity expansion in a system resource monitoring method based on a container cloud according to an embodiment of the present invention;

fig. 7 is a functional block diagram of a system resource monitoring apparatus based on a container cloud according to an embodiment of the present invention.

Detailed Description

The embodiment of the invention provides a system resource monitoring method, a system resource monitoring device, system resource monitoring equipment and a storage medium based on a container cloud, which are used for monitoring the resource use condition of a container cloud platform, early warning after problems are found in time and avoiding service paralysis caused by incapability of restarting of an application.

In order to make the technical field of the invention better understand the scheme of the invention, the embodiment of the invention will be described in conjunction with the attached drawings in the embodiment of the invention.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Fig. 1 is a flowchart of a system resource monitoring method based on a container cloud according to an embodiment of the present invention, and as shown in the drawing, the system resource monitoring method based on the container cloud includes steps S1 to S6:

and S1, generating a frame list after acquiring the deployment condition of the container deployment frame under the container cloud platform, and recording all the container deployment frames deployed under the container cloud platform in the frame list.

Specifically, a plurality of container arrangement tools are generally deployed on the container cloud platform through a container technology, and then various services or applications are allocated to corresponding system resources through a function cluster formed by the tools. The method comprises the steps of connecting to a management console of a container cloud platform by obtaining access authority of the container cloud platform, and then sending a data request command to the console to obtain deployment conditions. For example, the interface command "/ping" for acquiring the service state of the Marathon framework is used in the DCOS platform to call the running condition of Marathon. And summarizing all the acquired container arrangement frame information, and then generating a list or a name list according to the time of acquiring the corresponding container arrangement frame information. The list or the manifest is used as a location and a sequential reference for obtaining the running state of the application after being called by the subsequent steps.

And S2, acquiring the running state information of each application in each container arrangement frame one by one from the frame list according to a preset acquisition cycle according to the recording sequence, and recording the acquired running state information in a preset storage unit, wherein a frame recording node for recording the physical machine resource configuration data of each container arrangement frame and an application recording node for recording the running state information and the physical machine resource occupation data of each application are arranged in the storage unit, and the running state information is used for identifying the running state of the application in the container arrangement frame where the running state information is located.

Specifically, the information of each application managed by each container arrangement framework is obtained by running a command of each container arrangement framework, and then the information is stored in a recording node specially opened for the information, so that the data can be called in the subsequent step. For example, the requested content may be returned by calling a Marathon API interface and then sending a command to the management console of Marathon. For example, sending "/deployments" to the management console may obtain the deployment status of the applications on the current marathon orchestration framework, including the current resource occupancy status and the running status of each application. In addition, in the storage space where such recording nodes are arranged, corresponding recording nodes are opened up for the physical machine resource configuration data of the container arrangement framework, and the data in the recording nodes can be continuously and permanently stored according to the recording time sequence for being called by functional units for some analysis purposes, for example, in order to analyze the use condition of an application on a cloud platform in a certain period so as to deduce the development trend of the service corresponding to the application, the stored historical data is required to be used as a calculation basis.

And S3, when the running state information of any application is in a waiting state continuously within a preset judgment time threshold range, marking that the resources of the container arrangement frame are insufficient, and at the moment, generating alarm information and pushing the alarm information to an executor executing capacity expansion operation so as to inform the executor to execute the capacity expansion operation of the container arrangement frame.

Specifically, the temporary suspension or waiting state of some applications is not necessarily caused by resource allocation shortage, and the application is automatically restarted after a certain time, but if the application is waiting due to insufficient resources, the application is continued, and the application cannot be restarted, so that a determination time length needs to be preset, and if the state of a certain application is always waiting within the time length, it may be considered that the resource allocation condition of the application is insufficient to support the restart or normal operation of the application, and at this time, it may be considered that the resource of the container arrangement frame corresponding to the application is insufficient, and it is necessary to add sufficient hardware resources to the application, which is called capacity expansion. When the fact that resources of a certain container arrangement frame are insufficient is found, corresponding alarm information is generated and then pushed to an executor responsible for capacity expansion operation, such as a third-party maintenance company or a platform operation and maintenance company, and the pushing mode comprises mail, SMS message or voice dialing.

S4, acquiring the physical machine resource configuration data marked as the container arrangement frame with insufficient resources and the physical machine resource occupation data occupied by any application running in the container arrangement frame, and recording the two types of data in corresponding recording nodes.

Specifically, the management console of the container arrangement framework sends corresponding commands to acquire physical machine resource configuration data and physical machine resource occupation data, and then stores the two types of data in corresponding recording nodes.

And S5, after receiving the capacity expansion operation ending signal fed back by the executor, retrieving the previously recorded physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency from the storage unit, then reconfiguring and restarting the application, and recording the current physical machine resource configuration data marked as the container arrangement frame marked as resource deficiency in the frame recording node.

Specifically, the signal of the expansion ending operation may be acquired by setting a special feedback interface, and submitted after being input by the executor. An input entry for acquiring which newly added hardware resources are arranged on the feedback interface, so that an executor submits related information, and after the related information is acquired, the related information can be recorded in the frame recording node after the configuration data of the newly added physical machine resources are acquired. In addition, after the capacity expansion is finished, the application on the currently expanded container arrangement framework needs to be reconfigured and restarted, and the configuration basis is the latest recorded data which is stored in the application recording node before, including configuration data such as memory occupation and CPU thread allocation.

And S6, summarizing the recorded data of the framework recording node and the application recording node and then generating a capacity expansion report.

Specifically, after the capacity expansion is finished, in order to provide reference basis and data support for subsequent jobs, in addition to the backup data retained in the storage unit, the capacity expansion condition may be summarized and a job report may be generated, where data in the frame recording node and the application recording node before and after the capacity expansion is recorded.

In the embodiment, each container arrangement framework running in the container cloud platform is monitored, the running state of the application in the container arrangement framework is acquired, whether the situation of insufficient resources exists is judged, early warning is timely sent out, the application running is recovered after capacity expansion operation is completed, and service loss caused by single monitored nodes and incapability of timely early warning in traditional operation can be effectively avoided.

Fig. 2 is a flowchart of generating a frame list in the method for monitoring system resources based on a container cloud according to the embodiment of the present invention, where as shown in the drawing, in step S1, the frame list is generated after acquiring a deployment situation of a container arrangement frame in the container cloud platform, and all container arrangement frames deployed in the container cloud platform are recorded in the frame list, including step S101 to step S104:

and S101, connecting a management console of the container cloud platform.

S102, sending a data request for acquiring the condition of the container arrangement framework running on the container cloud platform to a management console of the container cloud platform.

Specifically, after the access right of the container cloud platform is acquired, the container cloud platform is connected to a management console, and then a data request is sent to request to acquire configuration data of a container arrangement framework deployed on the cloud platform, where the data request includes a command for acquiring the configuration data of the container arrangement framework. The management authority of the cloud platform comprises information such as an access address, a data port, a user name and a password.

S103, receiving the feedback of the management console, generating the frame list, and recording all container arrangement frames running on the container cloud platform in the frame list according to the feedback time sequence.

And S104, generating a recording sequence number for each container arrangement frame in the frame list according to recording time, wherein the recording sequence number is an identification sequence number of the container arrangement frame in a container cloud platform and is used for distinguishing different container arrangement frames.

Specifically, after data returned by the console are received, each acquired container arrangement frame is arranged into a list after being numbered according to the returned time sequence, and the subsequent steps are convenient to call and distinguish.

In the embodiment, all container arrangement frameworks running on the container cloud platform are arranged into the list, so that the calling of the subsequent steps is facilitated.

Fig. 3 is a flowchart of monitoring application states in the method for monitoring system resources based on a container cloud according to the embodiment of the present invention, where as shown in the figure, in step S2, the operation state information of each application in each container arrangement frame is acquired one by one from the frame list according to a preset acquisition cycle according to a recording sequence, and the acquired operation state information is recorded in a preset storage unit, where the method includes steps S201 to S204:

s201, generating a monitoring node for each container arrangement frame in the frame list, wherein the monitoring node is used for acquiring running state information of each application running on the monitoring node after the monitoring node is connected with a management console of the container arrangement frame in a set period.

Specifically, the monitoring node may be in a form of a function script composed of a command for accessing relevant information of the container arrangement frame, and the function script is configured to perform a data request on the relevant container arrangement frame at a specific time or in a specific period, so as to obtain an operation condition of an application thereon. The monitoring requirement of the framework can be arranged corresponding to a plurality of containers by setting the monitoring nodes.

S202, generating a corresponding application recording node for the application on each container arrangement frame in the storage unit according to the recording sequence number of the container arrangement frame in the frame list, wherein the application recording node is used for recording the running state information of each application running on the container arrangement frame, which is acquired by the monitoring node.

Specifically, in order to permanently store data such as application running conditions acquired by the monitoring node, the data may be recorded in a database or an independent data file, different recording nodes are set for the container arrangement framework itself and the application running on the container arrangement framework, and each type of recording node may sequentially record the acquired data according to the recording time as a sequence.

And S203, after the management control console of the container arrangement frame is connected according to a set monitoring period through the monitoring node, requesting to acquire the running state information of all the applications running on the container arrangement frame.

And S204, after receiving the feedback of the management console of the container arrangement framework, recording the running state information of the application in the application recording node according to the time when the feedback is received.

Specifically, a monitoring period is set for the monitoring node, which is equivalent to setting a function script with a monitoring function for executing a timing task, and a management console of a container editing framework and connection authority information of the recording node are configured in the script and have read-write authority. And after receiving the feedback data, writing the running data of the application in the application recording node by using the read-write permission, wherein the running data comprises the running state of the application. For example, states of Marathon managed applications include "waiting", "delaying", "suspending", "running". Wherein, waiting indicates that there is a case that some application or service is in failure or crash, and needs to restart the application or service; "latency" means that there is a delay in execution of an application or service due to resource exhaustion or congestion; "suspend" means that there is no application or service to be executed after being temporarily interrupted, and "run" means that the current application or service is in a normal running state. If an error is reported, this indicates that there is an application or service outage, and if such an event exists, in general, Marathon will throw a status word, such as "wait," to indicate that it is currently waiting for the relevant application or service to restart.

In the embodiment, the real-time application running condition is obtained by setting the monitoring node for the container arrangement framework, and the running data is permanently recorded for subsequent calling.

Fig. 4 is a flowchart illustrating a process of determining resource shortage in the container cloud-based system resource monitoring method according to the embodiment of the present invention, where as shown in the drawing, in S3, when the running state information of any one of the applications continues to be in a waiting state within a preset determination time threshold, the container arrangement frame is marked as resource shortage, and at this time, alarm information is generated and then pushed to an executor performing a capacity expansion operation, where the method includes steps S301 to S305:

s301, reading the running state information of any application in the application recording nodes.

S302, judging whether the running state information applied in the judging time threshold range is in a waiting state continuously, if so, marking the state of the container arranging frame as insufficient resources, and if not, marking the state of the container arranging frame as normal running, wherein the judging time threshold range is a preset period of time.

Specifically, if an application is in a waiting state within a set determination period, it may be considered that the application fails and needs to be rebuilt or restarted by the container arrangement framework, but when the application is always in the waiting state, it may be considered that the application cannot be recovered. Generally, an application running on a container orchestration framework is equivalent to an independent software program running in a virtual machine, and when the program crashes after being damaged or failed, generally, a virtual machine system will try to restart or wake the program, but for an application bound to a business relationship, the occupied resource of the application correspondingly changes along with the change of the business, and generally, the requirement on the resource is more and more without maintenance and optimization. When such a situation occurs, it is generally necessary to reconfigure resources, that is, expand hardware resources, of the container arrangement framework corresponding to the application, so as to allocate more hardware resources to the container arrangement framework for use, so that the container arrangement framework can allocate more resources to the application that has a problem, and the application can be re-created or restarted.

S303, traversing all applications under all container arrangement frames in the frame list according to the steps, and marking the states of all container arrangement frames.

Specifically, the serial numbers of the frames are arranged according to each container in the frame list, the running state data of the corresponding application is acquired from the application recording nodes one by one, whether the situation that capacity expansion is needed exists is judged, and then the judgment result is recorded.

And S304, generating an alarm mail after calling the mail template, and recording the recording sequence number of the container arrangement frame marked as the resource deficiency and the prompt information for marking the resource deficiency in the alarm mail.

S305, reading the mail address of the executor from a preset addressee address list and pushing the alarm mail to the executor.

Specifically, according to the record condition of the above steps, when a situation that resources are insufficient in a certain container layout frame occurs, an alarm mail with a specific format is generated by calling a mail template prepared in advance, in which the occurred problem and the location of the occurred problem are recorded, and then the alarm mail is sent to a handler according to mail address information, wherein the handler is generally an executor performing capacity expansion operation or a scheduling department, and the handler forwards the alarm mail to the executor. In addition, in other embodiments, the early warning effect may also be achieved by setting a voice dialing telephone with specific alarm content, for example, generating an early warning text according to the judgment record of the resource shortage condition, generating an early warning voice according to the text-to-speech translation engine, connecting with an executor, and playing the early warning voice. In some embodiments, the early warning information can be pushed in real time in a form of binding the mobile terminal APP of the invention and the executor.

In the embodiment, whether the container arrangement framework has the resource shortage condition or not is determined by judging the running state of the application, and the early warning mechanism is combined to realize the timely sending of the early warning information, so that the assistance is provided for timely meeting the capacity expansion requirement.

Fig. 5 is a flowchart of data backup before capacity expansion in the container cloud-based system resource monitoring method provided in the embodiment of the present invention, and as shown in the drawing, in S4, the physical machine resource configuration data marked as a container arrangement frame with insufficient resources and the physical machine resource occupation data occupied by any application running in the container arrangement frame are acquired, and the two types of data are recorded in corresponding recording nodes, including steps S401 to S403:

s401, connecting the management console of the container arrangement framework marked as resource deficiency.

S402, sending a data request to the management console, wherein the data request is used for acquiring the physical machine resource configuration data of the container arrangement frame and the physical machine resource occupation data occupied by any application running in the container arrangement frame.

Specifically, for capacity expansion preparation, the operation condition of the container arrangement framework with resource shortage needs to be recorded and stored in advance before capacity expansion, and the operation condition is reserved for recovery after subsequent capacity expansion. For this purpose, the configuration of the respective application is acquired and recorded. And sending a data request for acquiring the application state through a management console connected with the container arrangement framework, and acquiring corresponding data. The command for acquiring data is formulated according to the characteristics of each container arrangement frame. The deployment situation of the application corresponding to the ID can be obtained according to the ID of the application, for example, through a command of "/v 2/apps/{ ID }" or directly obtained after obtaining the application list, for example, through a command of "/v 2/groups/{ ID }" to obtain the application group identified by the ID.

And S403, after receiving the feedback of the management console, recording the physical machine resource configuration data in the frame recording node and recording the physical machine resource occupation data in the application recording node according to the time of receiving the feedback.

Specifically, after receiving the return of the console, the framework recording node and the application recording node are connected, and the two types of returned data records are stored.

In this embodiment, by recording the configuration condition of the application and the physical machine resource condition of the container arrangement framework before capacity expansion, a complete history data record is generated conveniently, and a data recovery basis is also provided for application recovery after capacity expansion.

Fig. 6 is a flowchart of resuming application operation after capacity expansion in the container cloud-based system resource monitoring method provided in the embodiment of the present invention, and as shown in the drawing, after receiving a capacity expansion operation end signal fed back by the executor, the S5 retrieves, from the storage unit, previously recorded physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency, performs reconfiguration and restart of the application, acquires current physical machine resource configuration data of the container arrangement frame marked as resource deficiency, and records the current physical machine resource configuration data in the frame recording node, where the steps of S501 —:

and S501, receiving feedback information of the executor, wherein the feedback information comprises a capacity expansion operation ending signal.

And S502, after the storage unit is connected, reading a record from the application recording node, wherein the record is the one of the physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency and is closest to the current time.

S503, configuring the corresponding application according to the physical machine resource occupation data, and restarting the application after configuration is completed.

Specifically, after the capacity expansion is finished, each application is restarted by acquiring the data backed up before the capacity expansion. The volume expansion end signal can be provided by the executor according to a preset input interface after the volume expansion operation is finished. When data is extracted from the recording nodes, the time of the latest recording from the current extraction time needs to be judged, and the extracted recording according to the time is backup data before capacity expansion of the current operation.

S504, a management console connected with the container arrangement frame acquires the current physical machine resource configuration data of the container arrangement frame, and records the acquired data in the frame recording node according to the acquisition time.

Specifically, after the application after capacity expansion is recovered, the current data of the physical machine resource configuration data is acquired by connecting the management console of the container arrangement frame, and the data is recorded into the frame recording node to generate a new data record, where the record before the record is the hardware resource allocation condition before capacity expansion and before allocation to the new hardware resource of the container arrangement frame. According to the physical machine resource configuration data in the framework record node, the relation between service development and hardware growth trend can be analyzed by matching with the recovery condition of the application.

In this embodiment, by calling the backup data of the application configuration condition before capacity expansion, application operation can be quickly restored after the capacity expansion is completed, and meanwhile, by recording the hardware change data of the container arrangement frame before and after capacity expansion, a data analysis basis can be provided for the functional department of business analysis.

In some embodiments, after generating the frame list after acquiring the container orchestration frame deployment situation under the container cloud platform, the method includes:

and connecting the management control console of each container arrangement frame one by one according to the recording sequence in the frame list. And arranging a frame for the container which is successfully connected, and generating a new recording sequence number after adding a success mark after the recording sequence number. And arranging a frame for the container with failed connection, adding a failure mark after the recording sequence number, and generating a new recording sequence number.

Specifically, in order to screen the recording sequence of the container arrangement frames in the frame list and improve the precision of connection access, connection confirmation can be performed on each container arrangement frame one by one according to the list sequence in advance, corresponding marks are generated according to the connection condition, and new recording sequence numbers are generated after the marks are attached to the recording sequence numbers of the container arrangement frames, so that the current connection state can be directly identified from the recording sequence numbers in the subsequent steps, and the current connection process can be skipped. Meanwhile, by adding the marks, a complete frame list can be reserved, and reconnection of the container arrangement frame with failed connection can be executed after the capacity expansion judgment of all container arrangement frames with successfully connected marks is executed, so that the judgment distortion caused by the capacity expansion of some container arrangement frames in the process of generating a new recording sequence number is avoided.

In some embodiments, the acquiring the running state information of each application in each container arrangement frame one by one from the frame list according to the recording order according to a preset acquisition period, and recording the acquired running state information in a preset storage unit includes:

and identifying the recording sequence number of the container arrangement frame, executing the operation of reading the running state information of the application in the container arrangement frame after the recording sequence number contains a success mark, and not executing the operation of reading the running state information of the application in the container arrangement frame when the recording sequence number contains a failure mark.

Specifically, through discerning new record serial number, can effectively avoid the container arrangement frame that the part has a problem, improve the accurate nature of connection location to improve the dilatation and judge the efficiency of operation.

In some embodiments, the present invention provides a system resource monitoring device based on a container cloud, as shown in fig. 7, including a list generating module, an application state obtaining module, an alarm information pushing module, a data recording module, an application restarting module, and a capacity expansion report generating module, where:

the list generation module 11 is configured to generate a frame list after acquiring a container arrangement frame deployment condition under the container cloud platform;

an application state acquiring module 12 configured to acquire the running state information of each application in each container arrangement frame one by one from the frame list according to a preset acquiring period according to a recording sequence, and record the acquired running state information in a preset storage unit;

the alarm information pushing module 13 is configured to mark that the container arrangement frame resources are insufficient when the running state information of any one of the applications is continuously in a waiting state within a preset judgment time threshold range, and push the running state information to an executor executing capacity expansion operation after alarm information is generated;

the data recording module 14 is configured to acquire physical machine resource configuration data of the container arrangement frame marked as resource deficiency and physical machine resource occupation data occupied by any application running in the container arrangement frame, and record the two types of data in corresponding recording nodes;

the application restarting module 15 is configured to receive the capacity expansion operation ending signal fed back by the executor, retrieve the previously recorded physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency from the storage unit, perform application reconfiguration and restart, acquire the current physical machine resource configuration data of the container arrangement frame marked as resource deficiency, and record the current physical machine resource configuration data in the frame recording node;

and the capacity expansion report generation module 16 is configured to generate a capacity expansion report after summarizing the recorded data of the frame recording nodes and the application recording nodes.

In some embodiments, the present invention provides a computer device, including a memory and a processor, where the memory stores computer readable instructions, and the computer readable instructions, when executed by the processor, implement the steps of the container cloud-based system resource monitoring method.

In some embodiments, the present invention provides a computer-readable storage medium, on which computer-readable instructions are stored, and the computer-readable instructions, when executed by one or more processors, implement the steps of the container cloud-based system resource monitoring method, wherein the storage medium may be a non-volatile storage medium.

Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by associated hardware instructed by a program, which may be stored in a computer-readable storage medium, and the storage medium may include: read Only Memory (ROM), Random Access Memory (RAM), magnetic or optical disks, and the like.

The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above-described embodiments are merely illustrative of some embodiments of the present application, which are described in more detail and detail, but are not to be construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims

1. A system resource monitoring method based on a container cloud is characterized by comprising the following steps:

acquiring the deployment condition of a container arrangement framework under a container cloud platform, and generating a framework list, wherein all container arrangement frameworks deployed under the container cloud platform are recorded in the framework list;

2. The method for monitoring system resources based on a container cloud according to claim 1, wherein a frame list is generated after the deployment condition of the container orchestration frame under the container cloud platform is obtained, and all container orchestration frames deployed under the container cloud platform are recorded in the frame list, and the method includes:

a management console connected to the container cloud platform;

3. The method for monitoring system resources based on container cloud according to claim 2, wherein the step of acquiring the running state information of each application in each container arrangement frame one by one from the frame list according to a recording sequence according to a preset acquisition period, and recording the acquired running state information in a preset storage unit comprises:

4. The method for monitoring system resources based on a container cloud according to claim 1 or 3, wherein when the running state information of any one of the applications continues to be in a waiting state within a preset judgment time threshold, it is marked that the resources of the container arrangement framework are insufficient, and at this time, after the alarm information is generated and pushed to an executor executing a capacity expansion operation, it is convenient to notify the executor to execute a capacity expansion operation of the container arrangement framework, including:

5. The method for monitoring system resources based on container cloud according to claim 1, wherein the acquiring physical machine resource configuration data of a container arrangement frame marked as resource deficiency and physical machine resource occupation data occupied by any application running in the container arrangement frame, and recording the two types of data in corresponding recording nodes, includes:

6. The method for monitoring system resources based on container cloud according to claim 1, wherein after receiving the expansion job end signal fed back by the executor, the method retrieves the previously recorded physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency from the storage unit, reconfigures and restarts the application, acquires the current physical machine resource configuration data of the container arrangement frame marked as resource deficiency, and records the current physical machine resource configuration data in the frame recording node, including:

7. The method for monitoring system resources based on container cloud according to claim 2, wherein after the framework list is generated after the container orchestration framework deployment condition under the container cloud platform is obtained, the method comprises:

8. A system resource monitoring device based on a container cloud is characterized by comprising:

the application restarting module is configured to receive a capacity expansion operation ending signal fed back by the executor, retrieve previously recorded physical machine resource occupation data occupied by each application in the container arrangement frame marked as resource deficiency from the storage unit, perform application reconfiguration and restart, acquire current physical machine resource configuration data of the container arrangement frame marked as resource deficiency, and record the current physical machine resource configuration data in a frame recording node;

and the capacity expansion report generation module is used for generating a capacity expansion report after summarizing the recorded data of the frame recording node and the application recording node.

9. A computer device comprising a memory and a processor, the memory having stored therein computer-readable instructions, wherein the computer-readable instructions, when executed by the processor, implement the container cloud based system resource monitoring method of any one of claims 1 to 7.

10. A computer-readable storage medium having computer-readable instructions stored thereon, wherein the computer-readable instructions, when executed by one or more processors, implement the container cloud based system resource monitoring method of any one of claims 1 to 7.