CN108153532A - A kind of cloud application dispositions method based on Web log mining - Google Patents
A kind of cloud application dispositions method based on Web log mining Download PDFInfo
- Publication number
- CN108153532A CN108153532A CN201711435285.1A CN201711435285A CN108153532A CN 108153532 A CN108153532 A CN 108153532A CN 201711435285 A CN201711435285 A CN 201711435285A CN 108153532 A CN108153532 A CN 108153532A
- Authority
- CN
- China
- Prior art keywords
- cloud application
- service
- deployment
- communication
- deployment strategy
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
-
- 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
- H04L67/34—Network arrangements or protocols for supporting network services or applications involving the movement of software or configuration parameters
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Invention is related to a kind of cloud application dispositions method based on Web log mining.Standardization defines communication log between cloud application first, is then based on data mining and daily record is analyzed, and is clustered according to the frequent degree to communicate between cloud application, finds out the cloud application set frequently to communicate, finally provide the prioritization scheme of cloud application deployment strategy accordingly;It realizes the collection and analysis of daily record data, deployment strategy prioritization scheme is provided according to the signal intelligence between service, and the self-adaption deployment for completing related cloud application is adjusted;Construct the adaptive cloud application frame of a performance oriented, including components such as gateway, log audit, resource management, container layout, service registration discovery and the automatic adjusuments of deployment strategy, it can not only realize the efficient deployment to cloud application and management, the deployment strategy of cloud application can also adaptively be adjusted according to the operating status of system, improve system performance.
Description
Technical field
The present invention relates to a kind of cloud application dispositions methods based on Web log mining, belong to software technology field.
Background technology
Tradition application is no longer complies with the growth requirement of new era, in order to adapt to big data, high concurrent, the business of mass memory
Demand, more and more new technologies, new concept are continuing to bring out, and wherein cloud computing, cloud application, containerization have been increasingly becoming one
The new trend of kind.Many Large-Scale Interconnected net companies have begun to recognize the various problems and limitation of singulation application, and open
Begin to use a kind of new software architecture pattern:Cloud application.Cloud application is a kind of lightweight software architecture mould based on SOA
Formula is applied to by Amazon, Nai Feideng companies in their system now.Cloud application emphasizes clear in job responsibility, promotion general's list
Body application is divided into one group of small service, is communicated between service using the communication mechanism of lightweight, mutually coordinated, is carried for user
For final value.Cloud application emphasizes the terseness of agile development, service logic and technology, allows development teams in the middle part of cloud environment
Administration's application, Quick Extended, are a kind of duration solutions towards cloud computing.
In terms of O&M is developed and singulation framework is different towards entire application, and cloud application focuses more on certain of system
One specific atomic function.The code library and O&M development teams that cloud application can have oneself independent are supported to use different technologies
Stack and database disclosure satisfy that system locally modification, newer requirement in Fast Iterative Procedure, while the lasting delivery of support,
The working efficiency of development teams can be greatly improved.In terms of scalability, cloud application is relatively independent, loose coupling between service,
It can independent, efficient, the flexible extension according to respective type of server in cloud environment.In terms of fault-tolerance, cloud application can
To avoid the single point failure problem of singulation framework, failure cloud application can be isolated, system is made to avoid losing because of a service
It imitates and causes entirely to apply the problem of delaying machine.Meanwhile by container technique, cloud application can realize the accurate control used resource
System improves resource utilization, supports quick, lightweight service startup and deployment.
It is a vital step in cloud application systematical design idea that service, which divides, is not only related to the exploitation of cloud application
Deployment, operation management, it is also closely bound up with cloud application system performance.Cloud application is operated in distributed environment, and quantity of service is many
More, operation management is complicated.Method call between each function module of monomer system is different, and lightweight is used between cloud application
Communication mechanism communicate, final value is provided to the user by mutual call coordinated between different services.Therefore, it is different
Service divide can generate different service call pattern and communication cost, Different Effects are also brought along to system performance.In order to
Better performance is obtained, in the design process of cloud application system, it would be desirable to the constantly partition strategy of modification and adjustment service.
One good cloud application division may mean that a successful cloud application system.But, it has been found that it is divided in cloud application
In the case of certain, the deployment strategy of service can also have an immense impact on to system performance.
Deployment strategy influences system performance.Cloud application framework is towards large-scale, complicated applications, and system is by large number of, mutual
Independent cloud application is formed.Compared with monomer applications, communication efficiency is low between cloud application, all can across main-machine communication, service call
Extend the response time of user's request, reduce throughput of system.Each cloud application of system, can portion there are many deployment selection
It affixes one's name on the arbitrary node in distributed environment, different deployment selections will generate different communication costs, such as:It will communicate frequent
Cloud application be deployed in same node and can reduce unnecessary across main-machine communication, communication cost can be reduced.Conversely, then will
Increase communication cost.In order to ensure system performance, we should as far as possible dispose close relation, the frequent cloud application of communication
Together, it avoids unnecessary across main-machine communication.Therefore, a deployment strategy packaged is of great significance to cloud application system.
Lack the dynamic mechanism of service arrangement.The deployment of cloud application mainly uses the static state based on service logic at present
Deployment strategy.On the one hand, the deployment strategy based on service logic can be close by logical relation or has the service department of business demand
Administration together, is reduced unnecessary across main-machine communication, raising system performance between servicing.On the other hand, cloud application is a kind of new
Emerging software design architecture, less to the system performance problems concern caused by deployment strategy in the industry at present, everybody is more concerned with cloud
Design, the service of application architecture divide and how to build cloud application frame etc..But it is found by experiment, based on service logic
Static deployment strategy be not ensure system performance optimal selection.It can be analyzed to a certain extent based on service logic
Signal intelligence between service, but be inaccurate.The operating status of system is the process of a dynamic change, logical between service
Letter situation is nor unalterable.It is unnecessary across main-machine communication between service in order to reduce to the greatest extent, to improve system performance, need
It builds real-time tracing and analyzes the mechanism that user asks service call relationship.By tracking and analysis, accurate acquisition service
Between signal intelligence, with this dynamic adjustment, Optimization deployment strategy.In addition to this, each cloud application possesses the process of oneself, itself
Can dynamic start and stop, for seamless upgrade and dynamic adjustment prepare.But who carrys out start and stop, which node what opportunity is selected at
This is that cloud application itself is uncontrollable for upper execution, and cloud application frame is needed to provide service management and deployment ability.
Invention content
The purpose of the present invention:Performance self-adapting adjusting method can realize the collection and analysis of daily record data, according to service
Between signal intelligence provide deployment strategy prioritization scheme, and complete related cloud application self-adaption deployment adjust.
The principle of the present invention:First standardization define cloud application between communication log, be then based on data mining to daily record into
Row analysis, is clustered according to the frequent degree to communicate between cloud application, finds out the cloud application set frequently to communicate, finally give accordingly
Go out the prioritization scheme of cloud application deployment strategy.
The technology of the present invention solution:A kind of cloud application dispositions method based on Web log mining, feature are to realize step
It is rapid as follows:
(1)Build cloud application Communication topology
Cloud application is operated in distributed environment, and quantity of service is numerous, and signal intelligence is not known between service, and a usual cloud should
All it is related with other several cloud applications, in system is really run, the call relation between service is intricate.It is inventing
In communication for service record cast based on daily record data, by defining cloud application Communication topology, shielding harness bottom is complicated
Signal intelligence, the communication details of specific service are isolated, provide the abstract view of clear close friend to the user.The structure is with oriented nothing
The formal definition of ring figure, is stored by adjacency matrix, supports the automatic conversion of pictograph, the topology that cloud application can communicate
Structure chain type is input in journal file.Request id therein is used for the user that present communications cloud application is marked to perform request pair
As.
The template of cloud application Communication topology mainly includes following sections:
L sources service:Service belonging to the service being currently invoked directly and model character.
L communication services:The service interacted with source service or other communication services.
L asks ID:ID is asked, the user for identifying the execution of present communications cloud application asks object.
L communication topology figures:The signal intelligence between source service and communication service is recorded in the form of adjacency matrix.
(2)Establish journal format template
Communication for service record cast based on daily record data is intended to through the signal intelligence between journal file record service, so
Journal format is of great significance to this model.Logging tools such as log4j, Blitz4j of comparative maturity etc. may be used in the industry at present
With efficiently, easily export more detailed system operation daily record, initialization procedure, service execution time including application program
Deng, but the daily record exported under default situations can not meet the needs of this model record communication for service.In new journal format template
Eliminate with the relevant unnecessary information of system, increase the terseness of daily record data.Five keywords of main definitions in template
Section, the signal intelligence between emphasis record service.After servicing start completion, each user's request call can all generate a day
Will records, and situation is called for recording the cloud application of current request, including time, interface and request id etc..In daily record template
Field name and meaning it is as follows:
L timestamps:It records current service to be called the time, service order during for labelled synthesis service chaining.
L system names:Record current application system name.
L api interface names:Record is currently performed the interface name of cloud application calling.
L request identifiers:Globally unique identifier records the id of current request promoter, is merging multiple journal files
When for identifying the multiple services for completing same request.
L cloud application names:The currently called cloud application name of record.
(3)Implement Partition of role
Communication for service record cast based on daily record data advocates Partition of role, by different role it is clear in job responsibility realize record
The efficient of the daily record data of signal intelligence collects and pre-processes between service.The model divides altogether two kinds of roles, log management section
Point and journaling agent node.For each cloud application operation node there are one journaling agent, which monitors day by particular port
Will leader information receives instruction, is periodically sent to daily record data.Log management node is responsible for log collection and pre- place
Reason.In addition, there is one layer of HTTP request header parsing interface in all roles of the model, it is responsible in parsing HTTP request head
Request identifier.
1)Log collection.Log management node in invention is worked based on configuration script, and log collection has initiative, pipe
It manages information of the node in configuration script and periodically pulls signal to the transmission daily record of journaling agent node, so the daily record in invention
Collection model, which uses, pulls model.Following information is mainly included in configuration script:
L attribute informations:Log management node self attributes information is described, including node address, communication port etc..
L agent node information:It describes service registration and finds central information, safeguard all cloud applications(Agent node)Address,
Thus management node inquires all agent node addresses.
The l communication informations:Described, daily record pulls signal the log management node collector journal period.
2)Log integrity.After log management node completes collection of log data, start to carry out all journal files pre-
Processing.It handles work and includes daily record merging and daily record cleaning
Log management node includes two components:Daily record combining block and daily record cleaning assembly.Daily record combining block receives all
The input of journal file, and analyze, completing each service call asked according to request identifier and timestamp reduction relies on
Relationship, while the Communication topology between cloud application is recorded in the form of adjacency matrix, after finally output merges
Journal file.
Daily record cleaning assembly is responsible for cleaning the journal file that previous step merges, and it is defeated to meet mining algorithm for finally output
Enter the data set of specification.Work is broadly divided into two parts:
L cleans invalid information:Invalid information refers to the unrelated information that communicates between service, including service initiation information, operation shape
State information, system error message and database information etc..
L extraction service chain informations:Start successfully from service, response user's request starts, and will complete the service each asked
Calling situation extracts.
Daily record cleaning carries out after daily record merges, the first user's request call recorded after starting successfully from cloud application
Place starts.When analyzing multiple journal files, record communication for service is found according to the daily record specification of invention definition and form first
Then daily record completes the service call chain specifically asked, and the call chain is proposed according to timestamp and request identifier synthesis
It takes out
(4)Mining Frequent communication service collection
By the processing of the communication for service record cast based on daily record data, the rule of signal intelligence between record cloud application can be obtained
Generalized data set, invention on the data set by performing Apriori algorithm Mining Frequent communication service collection.Frequent communication service
Collection is determined by support and confidence level this two evaluation indexes.The process of Mining Frequent communication service collection is exactly in Standardization Service
The process of support and confidence level that Apriori algorithm is calculated between all cloud applications is performed on chain data set.Mining process point
For three steps:1st, it initializes, generates frequent 1 item collection;2nd, it connects, the high frequent communication service of item is generated by the frequent communication service collection of low item
Collection;3rd, the frequent communication service collection of high item for being unsatisfactory for minimum support is deleted in beta pruning.Algorithm 2 illustrates the detailed of the mining process
Thin step.
(5)Generation deployment strategy optimization
There are a large amount of intersections between a series of frequent communication set that Apriori algorithm generates after performing, can not be by Apriori's
Implementing result is directly as deployment strategy prioritization scheme, it is also necessary to further analysis.Invention is to the Pruning strategy of Apriori algorithm
It is modified, further beta pruning has been done on the basis of original Result, allow its implementing result directly as deployment
The output content of strategy optimization.Further Pruning strategy:It is maximum to choose support counting in current frequently communication service collection S
Deployment schemes of the set of service max_set as a node, delete all set for having intersection with max_set in S, until
S is sky.The purpose of further beta pruning analysis is to find the frequent cloud application set of communication, provide deployment strategy optimization side
Case, and the communication not being related in 1 item collection that frequently communicates between cloud application, and using a cloud application as the deployment of a node
Scheme has little significance.
The present invention has the following advantages that compared with prior art:Propose a kind of cloud application portion analyzed based on daily record data
Policy optimization method is affixed one's name to, the signal intelligence in the form of daily record between record service analyzes daily record number by way of data mining
According to providing deployment strategy prioritization scheme.
Description of the drawings
Fig. 1 is cloud application service management 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 flow:
Other infrastructure component functions in system architecture are as follows respectively:
(1)Service registration finds center
Service registration finds that the status information and operating condition of registration and all cloud applications of maintenance system are responsible in center, including service
Name, IP address, port numbers, health status etc..Service registration in invention finds that center is realized by consul, using based on
The Socket files of Docker do service discovery, can be automatically performed the monitoring to adding in this node service state.It is inventing
Cloud application frame in service registration find center provide of both technical support:1)It serves as service registration and finds center.
Service registration finds that the mailing address and real-time servicing service state of all services, service are responsible in center in cloud application frame
Each service that provider reaches the standard grade is required for service registration to find center registration, for the inquiry of service call side, calls.2)
Service-seeking interface is provided.It needs to inquire address of service according to Service name when deployment is optimized to cloud application, to phase
The service of closing carries out start stop operation, and in service again deployment module, the api interface that we are exposed by Consul is realized to service
System optimization deployment is completed in inquiry.
Service registration finds that center is run in the form of Docker clusters, and Server and Client are disposed with Docker.
Service registration finds center by a Consul Server(If network congestion, Consul Server horizontal on demand can expand
Exhibition)It is formed with several Consul Client, Consul Server and Consul Client are operated in Docker.It is each
A service arrangement node(Physical machine/virtual machine)All it is a Consul Client, is responsible for reporting to Consul Server in real time
Accuse the service state on this node.And Consul Server are then responsible for summarizing all service status informations, provide inquiry to the user
Interface.The operation principle of Consul Server and Consul Client are as follows:
1)Consul Server:Each Consul Server is a Consul cluster, is arrived in order to real-time reception
Service status information on other nodes needs to configure Consul Server using node to be monitored as a Consul
Client is added in the cluster of Consul Server.
2)Consul Client:It is a Consul Client that a node is often added on Consul Server.
In order to which Consul Client is enable to report the service status information on this node to Consul Server in real time, it would be desirable to
Service discovery tool is installed on each Consul Client, the tool is by the way of the Socket files based on Docker
Service discovery is done, can be automatically performed the judgement to being added to this node service state.
(2)Service visualized management tool
Service visualized management tool is responsible for providing a friendly service management interactive interface to the user, allows user high
Effect easily realizes operation to resources such as service, container, mirror images.The tool is realized by Shipyard, with reference to container layout pipe
Science and engineering tool Docker Swarm can be managed collectively the service on multiple nodes.The tool is realized in a manner of visual
Management to related resource, including mirror image management, service management, node administration etc..
Service visualized management tool in invention cloud application frame mainly by Shipyard Controller and
RethinkDB is realized.Wherein RethinkDB is a database container, is used to store account, engine, service secret key, dilatation member
The information such as data.Shipyard Controller are Shipyard controllers, and realization and the Web for being responsible for Remote API are visual
Change the realization of operation interface.The tool is deployed in two Docker containers, is separately operable RethinkDB and Shipyard
Controller provides access entrance to the user by 80 ports.
(3)Monitoring resource component
Monitoring resource component is responsible for monitoring in real time the resource service condition of a node, such as CPU, memory.The component by
CAdvisor is realized, can not only monitor the resource service condition of cloud application on each node, moreover it is possible to the system of node host
Information is monitored.Meanwhile the component can also preserve, and pass through the information for monitoring resource by Influxdb
Grafana is shown.Monitoring resource component is made of three parts:CAdvisor, InfluxDB and Grafana.
1)cAdvisor.CAdvisor is the occupation condition that Google is used for analyzing running Docker containers
And the tool of performance characteristics.It collects, polymerize, handles and exports operation container by a running finger daemon
Relevant information, these information, which can include the other resource isolation parameter of container levels, the history behaviour in service of resource, reflection resource, to be made
With the block diagram with network statistical data complete history situation.Imctfy containers and Docker containers are supported at present.cAdvisor
It can join together to use with InfluxDB and Grafana, the two tools are time series respectively(time series)Number
According to library and the instrument board of index(metrics dashboard), store by them and show information.
2)InfluxDB.InfluxDB is a distributed sequential of increasing income, event and achievement data library.It is opened using Go language
Hair is relied on without outside, and design object is to realize the extension of distributed and horizontal extension, it has three big characteristics:1)Time
Series(Time series):It supports to use the function related with the time, such as maximum, minimum, summation;2)Metrics(Measurement):
Mass data can be calculated in real time;3)Events(Event):Support arbitrary event data.In monitoring resource component
In, InfluxDB be mainly used for storing user account information, on node host information and node all containers real time information.
3)Grafana.Grafana is the Open-Source Tools for being used to access InfluxDB developed by JavaScript, is provided
Multiple functional measurement instrument board and graphic editor support Graphite, InfluxDB and OpenTSDB.Grafana evidences
There is following characteristic:1)Available for visualization measurement data, the mode for providing powerful gracefulness goes to create, shares, browses data, instrument
Disk(dashboard)It is middle to show not homometric(al)(metric)Data in data source.2)Hot plug control panel is provided and can be expanded
The data source of exhibition can coordinate with other times sequence library and carry out display data.It is mainly used for information displaying in the assembly.
(4)Container layout management tool
Container layout management tool is responsible for completing the management to container cluster, can be completed in the case of not switching node to phase
Close the editing operation of container.The tool is realized by Docker Swarm, and container in cluster is opened by Docker API realizations
Stop, delete, check.
The tool is made of a Swarm Manager and several Swarm Node.One is run on Swarm Manager
Swarm Daemon, be responsible for front end communication and request scheduling, a Swarm Agent is run on each Swarm Node, is born
Duty communicates with Swarm Manager.When carrying out Docker cluster managements, user only needs to communicate with Swarm Manager, so
Swarm Manager select a Swarm Node operation container according to the information of Discovery Service afterwards.Wherein,
Swarm Daemon are a task dispatcher(Scheduler)And router(Router), itself does not run container, it is only
Receive the request that Docker Client are sended over, select the container of suitable Swarm Node management above, accordingly even when
Swarm Manager are because certain reason machines of delaying will not influence the container being currently running on other nodes.
Claims (1)
1. method characteristic is to realize that step is as follows:
Establish communication for service record cast:Journal format template is defined, realizes and same request service is completed in distributed environment
Synthesis;Division log handles role, improves log processing efficiency;Cloud application Communication topology is defined, shielding underlying services lead to
The complexity of letter;
Generate Optimization deployment strategy:Frequent communication service collection is excavated, and be further improved Apriori by Apriori algorithm
The Pruning strategy of algorithm enables directly output deployment strategy optimization;
Implement performance self-adapting dynamic regulation:According to the deployment strategy prioritization scheme provided, it is automatically performed the portion of related cloud application
Administration's adjustment, and ensure the consistency of gateway mapping address.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711435285.1A CN108153532A (en) | 2017-12-26 | 2017-12-26 | A kind of cloud application dispositions method based on Web log mining |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711435285.1A CN108153532A (en) | 2017-12-26 | 2017-12-26 | A kind of cloud application dispositions method based on Web log mining |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108153532A true CN108153532A (en) | 2018-06-12 |
Family
ID=62463161
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711435285.1A Pending CN108153532A (en) | 2017-12-26 | 2017-12-26 | A kind of cloud application dispositions method based on Web log mining |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108153532A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109376130A (en) * | 2018-10-18 | 2019-02-22 | 国云科技股份有限公司 | A method of the self-defined template based on cloudy platform records operation log |
CN110493657A (en) * | 2019-08-22 | 2019-11-22 | 兰州启源信息技术服务有限公司 | OnStream intelligent online steaming media platform system |
CN111459766A (en) * | 2019-11-14 | 2020-07-28 | 国网浙江省电力有限公司信息通信分公司 | Calling chain tracking and analyzing method for micro-service system |
CN113064872A (en) * | 2020-01-02 | 2021-07-02 | 武汉金山办公软件有限公司 | Log management method, device and system |
CN113760638A (en) * | 2020-10-15 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Log service method and device based on kubernets cluster |
CN115373699A (en) * | 2022-07-07 | 2022-11-22 | 北京三维天地科技股份有限公司 | Automated deployment method and system |
CN115967607A (en) * | 2022-12-25 | 2023-04-14 | 西安电子科技大学 | Template-based distributed internet big data acquisition system and method |
-
2017
- 2017-12-26 CN CN201711435285.1A patent/CN108153532A/en active Pending
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109376130A (en) * | 2018-10-18 | 2019-02-22 | 国云科技股份有限公司 | A method of the self-defined template based on cloudy platform records operation log |
CN110493657A (en) * | 2019-08-22 | 2019-11-22 | 兰州启源信息技术服务有限公司 | OnStream intelligent online steaming media platform system |
CN111459766A (en) * | 2019-11-14 | 2020-07-28 | 国网浙江省电力有限公司信息通信分公司 | Calling chain tracking and analyzing method for micro-service system |
CN111459766B (en) * | 2019-11-14 | 2024-01-12 | 国网浙江省电力有限公司信息通信分公司 | Micro-service system-oriented call chain tracking and analyzing method |
CN113064872A (en) * | 2020-01-02 | 2021-07-02 | 武汉金山办公软件有限公司 | Log management method, device and system |
CN113760638A (en) * | 2020-10-15 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Log service method and device based on kubernets cluster |
CN115373699A (en) * | 2022-07-07 | 2022-11-22 | 北京三维天地科技股份有限公司 | Automated deployment method and system |
CN115373699B (en) * | 2022-07-07 | 2023-03-31 | 北京三维天地科技股份有限公司 | Automated deployment method and system |
CN115967607A (en) * | 2022-12-25 | 2023-04-14 | 西安电子科技大学 | Template-based distributed internet big data acquisition system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108153532A (en) | A kind of cloud application dispositions method based on Web log mining | |
US20200334309A1 (en) | Triggering generation of an accelerated data model summary for a data model | |
US12003572B1 (en) | Two-way replication of search node configuration files using a mediator node | |
US20210042658A1 (en) | Facilitating concurrent forecasting of multiple time series | |
US8904341B2 (en) | Deriving grounded model of business process suitable for automatic deployment | |
US10095370B2 (en) | Network configuration and operation visualizing apparatus | |
US10992585B1 (en) | Unified network traffic controllers for multi-service environments | |
US10659609B2 (en) | Hierarchy based graphical user interface generation | |
US7734775B2 (en) | Method of semi-automatic data collection, data analysis, and model generation for the performance analysis of enterprise applications | |
US11172065B1 (en) | Monitoring framework | |
US20020103886A1 (en) | Non-local aggregation of system management data | |
CN111506412A (en) | Distributed asynchronous task construction and scheduling system and method based on Airflow | |
WO2020215324A1 (en) | Two-tier capacity planning | |
CN114693262A (en) | Smart city information grid operating system | |
WO2003034338A2 (en) | Management platform and environment | |
Graupner et al. | A framework for analyzing and organizing complex systems | |
US11934869B1 (en) | Enhancing efficiency of data collection using a discover process | |
Henning | Prototype of a scalable monitoring infrastructure for Industrial DevOps | |
JPWO2006051599A1 (en) | Resource management program, resource management method, and resource management apparatus | |
Safy et al. | Runtime monitoring of soa applications: Importance, implementations and challenges | |
EP1950909B1 (en) | Management system of a telecommunication network with a web-like graphic interface | |
CN113434268A (en) | Workflow distributed scheduling management system and method | |
Stanford | Geo-distributed stream processing | |
Florio | Design and management of distributed self-adaptive systems | |
US20230315580A1 (en) | Disaster recovery in a cell model for an extensibility platform |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180612 |