CN108182130A - A kind of cloud application container automatic monitoring method based on template - Google Patents

A kind of cloud application container automatic monitoring method based on template Download PDF

Info

Publication number
CN108182130A
CN108182130A CN201711313258.7A CN201711313258A CN108182130A CN 108182130 A CN108182130 A CN 108182130A CN 201711313258 A CN201711313258 A CN 201711313258A CN 108182130 A CN108182130 A CN 108182130A
Authority
CN
China
Prior art keywords
container
monitoring
data
monitoring data
template
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201711313258.7A
Other languages
Chinese (zh)
Inventor
周红卫
刘延新
李亚琼
周博
刘永波
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Run He Software Inc Co
Original Assignee
Jiangsu Run He Software Inc Co
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jiangsu Run He Software Inc Co filed Critical Jiangsu Run He Software Inc Co
Priority to CN201711313258.7A priority Critical patent/CN108182130A/en
Publication of CN108182130A publication Critical patent/CN108182130A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
    • G06F11/301—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system is a virtual computing platform, e.g. logically partitioned systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
    • G06F11/302—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system component is a software system
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3065—Monitoring arrangements determined by the means or processing involved in reporting the monitored data
    • G06F11/3068—Monitoring arrangements determined by the means or processing involved in reporting the monitored data where the reporting involves data format conversion
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3089—Monitoring arrangements determined by the means or processing involved in sensing the monitored data, e.g. interfaces, connectors, sensors, probes, agents
    • G06F11/3093—Configuration details thereof, e.g. installation, enabling, spatial arrangement of the probes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/32—Monitoring with visual or acoustical indication of the functioning of the machine
    • G06F11/323—Visualisation of programs or trace data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3466—Performance evaluation by tracing or monitoring
    • G06F11/3476—Data logging
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/12—Network monitoring probes
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
    • H04L67/025—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP] for remote control or remote monitoring of applications
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
    • H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
    • H04L69/161—Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields
    • H04L69/162—Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields involving adaptations of sockets based mechanisms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2201/00—Indexing scheme relating to error detection, to error correction, and to monitoring
    • G06F2201/815—Virtual

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Quality & Reliability (AREA)
  • General Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mathematical Physics (AREA)
  • Computer Hardware Design (AREA)
  • Computer Security & Cryptography (AREA)
  • Data Mining & Analysis (AREA)
  • Debugging And Monitoring (AREA)

Abstract

Invention is related to a kind of cloud application container automatic monitoring method based on template.To in container environment monitored target, collect Measure Indexes, collect monitoring data be described abstractly, for monitoring data acquire and processing model is provided.The monitoring data of container can be actively sent to monitoring data persistence module and abnormality detection module by the monitoring data collector and container self-discovery component run in physical machine.The essential information of container can be obtained automatically, detects the software environment of container, and matches corresponding monitoring standard module, so as to implement to monitor automatically when container instance generates.

Description

A kind of cloud application container automatic monitoring method based on template
Technical field
The present invention relates to a kind of cloud application container automatic monitoring methods based on template, belong to software technology field.
Background technology
Container is a kind of multiplexing operation system, provided for application process the virtualization technology of isolation mech isolation test (Xavier M G, Neves M V, Rossi F D, et al. Performance Evaluation of Container-Based Virtualization for High Performance Computing Environments[C]// 2013 21st Euromicro International Conference on Parallel, Distributed, and Network- Based Processing. IEEE Computer Society, 2013:233-240.).Due to being multiplexed in operating system Core so that container has the characteristics that resource overhead is small；By simulating sandbox environment, it ensure that container has very strong isolation. Container compared to other existing virtualization technologies, have many advantages, such as second grade create, deployment timeliness it is good (B. I. Ismail, E. Mostajeran Goortani, M. Bazli Ab Karim, et al. Evaluation of Docker as Edge computing platform [C] // Open Systems. IEEE, 2015.), therefore, Internet enterprises start to incline To in the demand for using extensive container cluster coping resources mutation.The characteristics of container, leads to the monitoring under container environment autonomous Property, be faced with new challenges in terms of accuracy and promptness.It is introduced separately below.
In terms of independence, Container Type is described using the software environment of container in invention, Container Type passes through container The middleware of middle deployment distinguishes, such as container common in operation system includes Tomcat containers, MySQL containers etc..By It constantly changes with business demand in the Container Type run in physical machine, next moment needs the Container Type monitored tool There is uncertainty.Meanwhile the light weight level characteristics of container so that thousands of a containers (Guedes E A can be generated in a physical machine C, Silva L E T, Maciel P R M. Performability analysis of I/O bound application on container-based server virtualization cluster[C]// Computers and Communication. IEEE, 2014:1-7.), so the number of containers monitored is needed to might have in different moments Very big variation.The reason of these two aspects, causes monitored container to have very strong dynamic.It can not determine when newly-increased quilt In the case of monitoring container and newly-increased monitored Container Type, traditional monitoring instrument.
In terms of accuracy, in container environment, the phenomenon that each container generated in physical machine is seized there are resource, when The resource availability of preceding container can be influenced by the resource usage amount of other containers.For example, at monitoring initial stage, monitoring system is only It needs to monitor Tomcat containers, in traditional monitoring mode, monitoring system can be that the measurement setting of Tomcat containers is fixed Threshold value, such as its memory is set as 1GB using threshold value, alarm operation is triggered when memory usage amount is more than 1GB.A certain Time point has increased a MySQL container newly in system, due to there are problems that resource is seized in container environment, so Tomcat is former Some memory thresholds will be adjusted according to MySQL free memory amounts, be otherwise easy to that memory overflow exception occurs, supervised When xylometer quantity size is very big, the adjustment of this threshold value becomes unrealistic.Therefore, discomfort degree of being combined under container environment Amount setting static threshold carries out intensity of anomaly assessment.
In terms of promptness, how long monitoring system is it can be found that exception, is largely dependent upon monitoring cycle.Monitor work Tool needs to weigh between monitoring promptness and expense, because if monitoring cycle is set due to using fixed monitoring cycle The too small then expense put is larger, and monitoring cycle is excessive and may fail to report exception.Meanwhile existing research has shown that node is on a grand scale Cluster monitoring can occupy the system resources such as more network bandwidth because clustered node is likely located at the different number in geographical location According to center (Raghavan B, Vishwanath K, Ramabhadran S, et al. Cloud control with distributed rate limiting[J]. Acm Sigcomm Computer Communication Review, 2007, 37(4):337-348.).Therefore, it is difficult that certain monitoring is being met in time by fixed monitoring cycle under container environment It is monitored under the premise of property with relatively low expense.
Container monitoring problems faced concentrates on three aspects：The characteristics of for monitored container dynamic change, how to adopt Newly-generated container is found in time with effective method and detects the type of container, so as to be collected into current container physics The measurement of resource dimension and logical resource dimension；Using more accurately method to current in the container environment how to be seized in resource Container carries out intensity of anomaly assessment；Effective algorithm dynamic adjustment monitoring cycle how to be used to save expense.
Invention content
The purpose of the present invention：The generation of container is perceived in time and detects the software environment of container, according to the software of container Environment binds corresponding monitoring standard module, and the monitoring data of container are then non-invasively obtained in physical machine, can be found automatically Container.
The principle of the present invention：The monitoring data collector and container self-discovery component run in physical machine can be by the prison of container Measured data is actively sent to monitoring data persistence module and abnormality detection module.Abnormality detection module is according to the monitoring being collected into Then intensity of anomaly is passed to monitoring cycle adjustment module by the intensity of anomaly of data assessment current container in the form of a signal. Transmission frequency of the monitoring cycle adjustment module according to the adjustment data collection instruction of intensity of anomaly signal.
The technology of the present invention solution：A kind of cloud application container automatic monitoring method based on template, feature are Realize that step is as follows：
The first step establishes monitoring model：
Define 1：Host model：Host: = (Specification, Instances)
Specification:=(name, IP, username, password, docker_port)：Place where representing container The detailed description of host,nameRepresent host title,IPRepresent the IP address of host,usernameWithpasswordRespectively Represent username and password,docker_portRepresent Docker Daemon port numbers open on current host.
Instances:={Instance i }：Represent all container instances run on current host.
Define 2：Container model：Instance: = (Specification, DataSet, Metrics)
Specification:= (uuid, name, type, host, port)：Represent the detailed description of container instance,uuidRepresent container instance unique mark,nameRepresent the title of container instance,typeRepresent the software type in container,host Represent the affiliated host IP address of container,portRepresent container mappings to the port numbers of host.
DataSet:={Data i }：Monitoring data set on current container.
Metrics:={Metric i }：The Measure Indexes set that current container needs are collected.
Define 3：Metric index model：Metric: = (Instance, Group, Unit)
Instance：Container instance belonging to current metric index.
Group：Grouping belonging to current metric index, such as CPU, memory, Tomcat processing number of request.
Unit：The unit of current metric index, such as MB, KB/s etc..
Define 4：Monitoring data model：Data: = (Instance, Metric, time, value)
Instance：Container instance belonging to the currently monitored data.
Metric：The corresponding Measure Indexes of the currently monitored data.
time：The timestamp of the currently monitored data.
value：The monitoring data value being collected into.
It is the abstractdesription to monitored target to define 1 and define 2, and defining 3 and defining 4 pairs needs the monitoring data collected Information is described.The data gathering form defined by above-mentioned model can be represented effectively distributed, unstructured, different The monitoring data of structure.
Since the Measure Indexes that same type of container needs monitor are relatively fixed, this section is based on defined above, pumping Model as having gone out monitoring standard module, is defined as follows：
Define 5：Monitoring standard module model：Template: = (Specification, Metrics, graph)
Specification:=(uuid, name, type)：Define the unique mark of monitoring standard module, the name of template, mould The corresponding Container Type of plate.
Metrics：Measure Indexes set defined in current template.
graph：The figure of monitoring data displaying.
Second step, container are found automatically：The automatic perception for finding not only to have included container life cycle state, but also including container software ring The middleware example that the container software environment mentioned refers in particular to dispose in container is invented in the detection in border（Such as Tomcat, MySQL, Nginx etc.）.After perceiving the generation of container and detecting the software environment of container, need to need what is monitored for container binding Measurement, for this purpose, invention defines different monitoring standard modules for each different types of container.Monitoring standard module is with the shape of XML file Formula exists, and the group where template is contained in the definition of monitoring standard module, needs the measurement monitored, the figure of monitoring data displaying The information such as the trigger that shape and monitoring, alarming are used.
In order to obtain the information of container in a manner of non-intruding, information of container Register is read from configuration file first The information of Docker container hosts is taken, these information include that the IP address of host, Docker Daemon are opened on host Then the http port number put establishes the connection with Docker Daemon to obtain the relevant information of Docker, these information are made Input for algorithm.If successfully obtaining the UUID information of container, information of container Register just passes it to container Software environment detector.
Container software environmental sensor is mainly used to the container UUID information obtained according to information of container collector to detect The software environment of current container, the software environment of container refer in particular to the middleware disposed in container（Such as Tomcat, MySQL, Nginx etc.）Type.Container software environmental sensor is according to the process number of the UUID acquisition of information current containers of container, Ran Houjin Enter the corresponding name space of the process number.Container software environment detection order, the output of probe command are performed in name space As a result it can go in container software environment dictionary to inquire as index value, the software environment of container is finally determined according to query result.
Monitoring system defines monitoring standard module for the common middleware of industry, and template definition needs the measurement monitored, no The container of same type corresponds to different monitoring standard modules.Monitoring object matching module is container according to the container software environment detected Suitable monitoring standard module is matched, the IP address of monitoring system web front end module, user name, close is then read from configuration file The information such as code establish the connection with web front end module, will be sent to web front end module for the request of container addition monitoring, in this way It is just added to CPU usage, memory usage amount, network speed, magnetic automatically for monitored container in the web front end module of monitoring system The monitored item such as monitoring information of disk I/O speed, middleware, while the figure of data displaying can also be automatically created.
Third walks, and container monitoring data are collected.Intrusive monitoring mode needs to dispose monitor agent program in container For collecting data, this has violated the light weight level characteristics of container.Simultaneously as container operates in host in the form of process On, the monitoring data that the monitor agent program inside container is collected into are the monitoring data of host in fact.More than being based on Consider, invent the monitoring system of design using the monitoring mode of non-intruding the deployment container monitoring data collection group on host Part.Information of container is completed after the registration of monitoring system by the automatic discovery technique of container, monitoring system will start container prison Measured data collection assembly.Container monitoring data collection assembly in monitoring system is responsible for the monitoring data of collection vessel, these prisons Measured data had both included the measurement of container physical resource dimension, such as memory usage amount, CPU usage, network throughput and disk Rate etc. is read and write, and includes the measurement of container logical resource dimension, such as Tomcat is in Thread Count, the MySQL of busy condition Perform database insert operation number, the measurement in terms of Nginx actively connects the middlewares such as number.
Intrusive monitoring mode needs to dispose monitor agent program in container for collecting data, this has violated container Light weight level characteristics.Simultaneously as container is operated in the form of process on host, the monitor agent journey inside container The monitoring data that sequence is collected into are the monitoring data of host in fact.Based on considerations above, the monitoring system for inventing design uses The monitoring mode of non-intruding deployment container monitoring data collection assembly on host.It completes to hold by the automatic discovery technique of container For device information after the registration of monitoring system, monitoring system will start container monitoring data collection assembly, the framework of the component.Prison Container monitoring data collection assembly in examining system is responsible for the monitoring data of collection vessel, these monitoring data both include container object The measurement of resource dimension, such as memory usage amount, CPU usage, network throughput and disk read-write rate etc. are managed, and including holding The measurement of device logical resource dimension, such as Tomcat is in the Thread Count of busy condition, MySQL performs database insert operation Number, Nginx actively connect the middlewares such as number in terms of measurement.
1）The Metric Data Collection of physical resource dimension
Container physical resource dimension data collector is mainly connected by establishing TCP from the Docker Daemon in different physical machines Fetch acquisition data.Since the monitoring data that each container is obtained by Docker Daemon can consume the specific time, Container physical resource dimension data collector non-invasively obtains in a manner that multi-threaded parallel collects data and is located at not jljl Container monitoring data on reason machine, this data collection mode is compared with serially collecting the mode of data, in number of containers scale Data collection time can effectively be shortened when larger.Container finds that component believes the container for needing to monitor substantially automatically Breath is registered to monitoring system, the Docker Daemon that these information are configured including physical machine IP address, the physical machine where container Remotely access the port numbers used, the UUID of container, container name etc..Container physical resource dimension data collector is according to prison The physical machine IP address and Docker Daemon distance access ports slogans that examining system is capable of providing, pass through network communication components It establishes and is connected from the TCP communication of Docker Daemon in different physical machines, is successfully established after connection, data collector can be Each container in operating status is opened a thread and is collected for the monitoring data of physical resource dimension.Container physical resource Dimension data collector obtains the monitoring data JSON format strings of current container, character using the UUID of container as index value The metric data of the physical resources dimensions such as CPU, memory, magnetic disc i/o, network is included in string.
2）The Metric Data Collection of logical resource dimension
Container logical resource dimension data collector can obtain the measurement of container logical resource dimension, these are measured in invention Refer to the middleware disposed in container（Such as Tomcat, MySQL, Nginx etc.）Measure Indexes.
By the investigation for industry mainstream middleware monitoring mode, the monitoring agreement that invention sums up middleware can divide The type different for three kinds, including the monitoring based on JMX, the monitoring based on Socket and the monitoring based on HTTP, the table from Whether whether monitoring agreement need to change monitored target configuration file and self-defined monitoring script needed to parse data three Aspect compares different monitoring modes, and finally having been illustrated that each monitoring mode is corresponding can monitoring object.
According to the container software environment that the automatic discovery module of container detects, monitoring agreement matching component can be container logic Resource dimension data collector matches corresponding monitoring method.The data collection mode of each method is described in detail below.
Monitoring based on Socket agreements：This monitoring mode is introduced in invention by taking MySQL as an example.Built-in pipe in MySQL Science and engineering has mysqladmin, is transported in the IP address of host, monitored container where monitored container can be specified by parameter Capable MySQL is mapped to the port numbers of host, needs the information such as the MySQL obtained measurements.Container logical resource dimension data Collector performs corresponding mysqladmin orders in the form of parameter is transmitted, then the implementing result of resolve command.Pass through this Kind mode, can obtain following types of MySQL monitoring data：Change database number, perform Select operation number, The number for performing Insert operations, the number for performing Update operations, the number for performing Delete operations, maximum allowable connection Number, time for practical maximum number of connections, currently connecting number, actively connecting number, caching connection number, index reading times, caching query Byte number that the byte number and MySQL that number, MySQL are sent receive etc..
Monitoring based on JMX agreements：This monitoring mode needs to configure monitored middleware to open the branch to JMX agreements It holds.The monitoring based on JMX is introduced in invention by taking Tomcat as an example.Tomcat in operating status, can as Mbean Server To receive the data inquiry request from JMX clients at any time.Container logical resource dimension data collector is by monitoring system The IP address of host, the port numbers of the JMX port mappings of monitored container to host where obtaining monitored container, then The connection with Mbean Server is established, the request expression formula that input meets JMX protocol specifications can obtain monitored container Corresponding measurement.In this way, monitoring system can obtain the following types of monitoring data of Tomcat：Tomcat maximum threads When number, Tomcat current threads number, Tomcat current business Thread Count, Tomcat maximum processing time, Tomcat average treatments Between, Tomcat number of requests per second, the number of request of Tomcat failures per second, Tomcat send byte number, Tomcat receive word Memory, JVM can be used to use memory etc. for joint number, JVM.
Monitoring based on http protocol：Part middleware provides third-party monitoring modular, is opened by these modules API, can remotely obtain the monitoring data of middleware.The monitoring based on API is introduced in invention by taking Nginx as an example.Pacify in compiling It when filling Nginx, needs to enable http_stub_status_module modules, then specifies and obtain in Nginx configuration files Take the API that monitoring data use.Container logical resource dimension data collector obtains monitored container place by monitoring system The IP address of host, monitored container Nginx be mapped to the port numbers of host, then by IP address, port numbers and API is packaged into the host that a HTTP request is sent to where monitored container.Host returns to monitoring data response character After string, container logical resource dimension data collector therefrom parses the data of needs.In this way, monitoring system can To obtain following types of Nginx monitoring data：Actively connect number, it is processed connection number, have built up number of shaking hands, Processed number of request, etc. pending number of request, read client Header Information Numbers, return to client Header Information Numbers.
After the metric data of two dimensions of container physical resource and logical resource is all collected into, container monitoring data envelope Arrangement can carry out three kinds of processing to the data being collected into：1）The data that can be handled by MySQL persistences are parsed into, data are deposited Enter database；2）The character string of JSON forms is packaged into, is sent to the web front end module of monitoring system, the module is by data solution After analysis, the displaying of data is carried out on interface；3）Vector is parsed into, passes to the intensity of anomaly evaluation module of monitoring system, With reference to the unusual condition of historical data assessment current container.
The present invention has the following advantages that compared with prior art：
（1）The characteristics of for big number of containers scale, type dynamic change is monitored in container environment, using container self-discovery group Part and monitoring data collector non-invasively complete the registration of container essential information and monitoring number in the physical machine where container According to collection；
（2）Being seized for resource in container environment causes to be difficult to the characteristics of given threshold detection is abnormal, is examined using independent exception Module is surveyed, carrying out online intensity of anomaly to the monitoring data being collected into assesses；
（3）For container environment expense it is sensitive the characteristics of, by the way of actively monitoring data are sent, i.e., monitored node is being supervised After measured data collects completion, monitoring data are initiatively sent to monitoring data persistence module.
Description of the drawings
Fig. 1 is monitoring system software framework.
Specific embodiment
Below in conjunction with specific embodiments and the drawings, the present invention is described in detail as shown in Figure 1, embodiment of the present invention side Method and monitoring system software framework include following functions module：
Container self-discovery module：The information and Docker of Docker container hosts are read first from configuration file Then the http port number that Daemon is opened establishes the connection with Docker Daemon to obtain the relevant information of Docker, this A little information exist in the form of JSON format strings.After success obtains the UUID information of container, it can be obtained according to the UUID The process number of current container is taken, subsequently into the corresponding name space of the process number.Container software ring is performed in name space Border probe command, the output result of probe command can go in container software environment dictionary to inquire as index value, and last basis is looked into The software environment that result determines container is ask, suitable monitoring standard module, in this way the web front end module in monitoring system are matched for container In be just added to centre in CPU usage, network speed, magnetic disc i/o speed, memory usage amount, container automatically for monitored container The monitored item such as monitoring information that part needs, at the same the figure of data displaying can also be automatically created, abnormal alarm uses touches Send out device.
Monitoring data collection module：It is responsible for the monitoring data of collection vessel, these monitoring data had both included container physics and provided The measurement of source dimension, such as CPU usage, memory usage amount, disk read-write rate, network throughput isometry, and including holding The measurement of device logical resource dimension, such as Tomcat is in the Thread Count of busy condition, MySQL performs database insert operation Number, Nginx actively connect the middlewares such as number in terms of measurement.In addition, monitoring data collection module has non-intruding, multithreading simultaneously The characteristics of row is collected, actively sent, can effectively shorten data collection time.
Monitoring data persistence module：The time interval of monitoring data persistence can be configured in module, which can also The essential information of host, the title of monitored container, the software environment title of container, alarm phase where the monitored container of storage Measurement is monitored in the information of pass, container<Key, value>Etc. information.
Template group, item to be monitored, the figure of monitoring data displaying and alarm are included in the definition of monitoring standard module The information such as trigger, monitoring standard module exist in the form of an xml-file.
Information of container Register reads the information of Docker container hosts, Docker first from configuration file Then the http port number that Daemon is opened establishes the connection with Docker Daemon to obtain the relevant information of Docker, this A little information exist in the form of JSON format strings.Meanwhile that the UUID information of container is passed to container is soft for container Register Part environmental sensor.
Container software environmental sensor is currently held using docker inspect orders according to the UUID acquisition of information of container The process number of device, subsequently into the corresponding name space of the process number.Container software environment detection life is performed in name space Enable such as ps aux | grep mysql.Mentioned order is performed, and obtain output result using the commands modules of python As the foundation with the presence or absence of certain type of intermediate member makes.
Monitoring object matching module matches corresponding monitoring standard module according to the container software environment detected for container, from The information such as IP address, user name, the password that monitoring system is read in file are put, the connection with monitoring system is established, will be container The request for adding monitoring is sent to monitoring system, can be just monitored container addition CPU usage in this way in monitoring system, interior The monitored items such as usage amount, network speed, magnetic disc i/o speed, the relevant monitoring information of middleware are deposited, while data can also be automatically created The figure of displaying, the trigger that abnormal alarm uses.
The component registration of monitoring system reads UUID list objects, and all containers then are registered to monitoring system, adds Item is monitored, creates trigger.

Claims (1)

1. method characteristic is to realize that step is as follows：
The first step establishes monitoring model：Including host model, metric index model, monitoring data model, defined by model Data gathering form, can effectively represent distributed, unstructured, isomery monitoring data, monitoring standard module model：Template:=(Specification, Metrics, graph), Specification:=(uuid, name, type)：The unique mark of monitoring standard module, the name of template, the corresponding Container Type of template are defined,Metrics：Current template Defined in Measure Indexes set,graph：The figure of monitoring data displaying；
Second step, container are found automatically：The automatic perception for finding both to have included container life cycle state, the spy of container software environment It surveys, after perceiving the generation of container and detecting the software environment of container, needs to need the measurement monitored for container binding, be Each different types of container defines different monitoring standard modules；
Third walks, and container monitoring data are collected：Registration of the information of container in monitoring system is completed by the automatic discovery technique of container After, monitoring system will start container monitoring data collection assembly, and monitor agent program is disposed in container for collecting data.
CN201711313258.7A 2017-12-12 2017-12-12 A kind of cloud application container automatic monitoring method based on template Pending CN108182130A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711313258.7A CN108182130A (en) 2017-12-12 2017-12-12 A kind of cloud application container automatic monitoring method based on template

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711313258.7A CN108182130A (en) 2017-12-12 2017-12-12 A kind of cloud application container automatic monitoring method based on template

Publications (1)

Publication Number Publication Date
CN108182130A true CN108182130A (en) 2018-06-19

Family

ID=62546023

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711313258.7A Pending CN108182130A (en) 2017-12-12 2017-12-12 A kind of cloud application container automatic monitoring method based on template

Country Status (1)

Country Link
CN (1) CN108182130A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109101408A (en) * 2018-07-20 2018-12-28 阿里巴巴集团控股有限公司 The detection method and device of service availability in joint debugging environment
CN110752964A (en) * 2019-09-06 2020-02-04 锐捷网络股份有限公司 Network equipment testing method and device
CN111240663A (en) * 2019-12-30 2020-06-05 中国建设银行股份有限公司 Method and device for automatically building host middleware CICSPLEX system
WO2020125265A1 (en) * 2018-12-21 2020-06-25 中兴通讯股份有限公司 Container service monitoring method, system and computer readable storage medium
EP3640803A4 (en) * 2018-08-15 2020-09-02 Wangsu Science & Technology Co., Ltd. Host monitoring method and device
CN112199247A (en) * 2019-07-08 2021-01-08 中国移动通信集团浙江有限公司 Method and device for checking Docker container process activity in non-service state

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109101408A (en) * 2018-07-20 2018-12-28 阿里巴巴集团控股有限公司 The detection method and device of service availability in joint debugging environment
CN109101408B (en) * 2018-07-20 2021-09-07 创新先进技术有限公司 Method and device for detecting service availability in joint debugging environment
EP3640803A4 (en) * 2018-08-15 2020-09-02 Wangsu Science & Technology Co., Ltd. Host monitoring method and device
WO2020125265A1 (en) * 2018-12-21 2020-06-25 中兴通讯股份有限公司 Container service monitoring method, system and computer readable storage medium
CN111355622A (en) * 2018-12-21 2020-06-30 中兴通讯股份有限公司 Container traffic monitoring method, system and computer readable storage medium
CN112199247A (en) * 2019-07-08 2021-01-08 中国移动通信集团浙江有限公司 Method and device for checking Docker container process activity in non-service state
CN112199247B (en) * 2019-07-08 2022-07-01 中国移动通信集团浙江有限公司 Method and device for checking Docker container process activity in non-service state
CN110752964A (en) * 2019-09-06 2020-02-04 锐捷网络股份有限公司 Network equipment testing method and device
CN111240663A (en) * 2019-12-30 2020-06-05 中国建设银行股份有限公司 Method and device for automatically building host middleware CICSPLEX system

Similar Documents

Publication Publication Date Title
CN108182130A (en) A kind of cloud application container automatic monitoring method based on template
CN108475360B (en) Distributed computing dependency management system
US10089384B2 (en) Machine learning-derived universal connector
JP5471859B2 (en) Analysis program, analysis method, and analysis apparatus
US10225375B2 (en) Networked device management data collection
US8041683B1 (en) Methods and apparatus for locating network logs
US20230040635A1 (en) Graph-based impact analysis of misconfigured or compromised cloud resources
US11977934B2 (en) Automation solutions for event logging and debugging on KUBERNETES
CN102929613B (en) The tuning apparatus and method of operating system
CN103150250B (en) Application program capacity detection system and application program capacity detection method
CN103986625A (en) Cloud application fault diagnosis system based on statistical monitoring
CN105122733B (en) Queue is monitored and visualized
US11068300B2 (en) Cross-domain transaction contextualization of event information
CN108108288A (en) A kind of daily record data analytic method, device and equipment
JP2002342128A (en) Method to extract health of service from host machine
JP2023542646A (en) Smart span prioritization based on plug-in service backpressure
CN110546615A (en) Hyperdynamic JAVA management extensions
CN111654495B (en) Method, apparatus, device and storage medium for determining traffic generation source
US9397921B2 (en) Method and system for signal categorization for monitoring and detecting health changes in a database system
CN110166440A (en) Print protocol vulnerability analysis method and system
JP3897897B2 (en) TROUBLESHOOTING DEVICE AND TROUBLESHOOTING METHOD IN A NETWORK COMPUTING ENVIRONMENT AND COMPUTER-READABLE RECORDING MEDIUM CONTAINING TROUBLESHOOTING PROGRAM
US10848371B2 (en) User interface for an application performance management system
US11757736B2 (en) Prescriptive analytics for network services
CN111177100A (en) Training data processing method and device and storage medium
Kim et al. Optimizing Logging and Monitoring in Heterogeneous Cloud Environments for IoT and Edge Applications

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20180619

WD01 Invention patent application deemed withdrawn after publication