CN112685246B - Time sequence data processing method and device - Google Patents
Time sequence data processing method and device Download PDFInfo
- Publication number
- CN112685246B CN112685246B CN202011539336.7A CN202011539336A CN112685246B CN 112685246 B CN112685246 B CN 112685246B CN 202011539336 A CN202011539336 A CN 202011539336A CN 112685246 B CN112685246 B CN 112685246B
- Authority
- CN
- China
- Prior art keywords
- data
- time sequence
- sequence data
- type
- time
- 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.)
- Active
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Debugging And Monitoring (AREA)
Abstract
The invention discloses a time sequence data processing method and device, comprising the following steps: determining first time sequence data of a first computing node; and second time sequence data of a second computing node matched with the first computing node; determining a transition position in the first time sequence data and the second time sequence data if the first time sequence data and the second time sequence data meet corresponding morphological relations; determining first-type state data in the first time sequence data and first-type state data in the second time sequence data according to the conversion position; determining third time sequence data according to the first type of state data in the first time sequence data and the first type of state data in the second time sequence data; the problem that the load condition of the computing node can only be reflected by the time sequence data in the prior art and the integral load condition of the HA architecture cannot be reflected is avoided, and various analysis and extension applications are facilitated to the integral load condition of the HA architecture.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method and an apparatus for processing time sequence data.
Background
A dual-computer cluster system, or high availability architecture (HA for short), is a solution in the database/server field that can effectively guarantee service continuity. The architecture is characterized by generally comprising 2 computing nodes, wherein one computing node bears task load, and the other computing node is used as a standby. If necessary, 2 nodes can be switched, i.e. the task load is quickly transferred to the standby computing node. Thereby ensuring the safety and continuity of the system operation.
In this scenario, each compute node in the HA architecture HAs its own timing data. The time series data may generally reflect the two-dimensional relationship of load/time on the compute node, as shown in fig. 1. However, the time sequence data in fig. 1 can only represent the load condition of the computing node itself, but cannot reflect the actual effective load condition of the HA architecture, so that it is not beneficial to perform various analyses and extension applications on the actual effective load condition of the HA architecture.
Disclosure of Invention
The invention provides a time sequence data processing method and device, which at least solve the technical problems in the prior art.
In a first aspect, the present invention provides a method for processing time series data, including:
determining first time sequence data of a first computing node; and second time sequence data of a second computing node matched with the first computing node;
determining a transition position in the first time sequence data and the second time sequence data if the first time sequence data and the second time sequence data meet corresponding morphological relations;
determining first-type state data in the first time sequence data and first-type state data in the second time sequence data according to the conversion position;
and determining third time sequence data according to the first type of state data in the first time sequence data and the first type of state data in the second time sequence data.
Preferably, the determining the transition positions in the first time sequence data and the second time sequence data if the first time sequence data and the second time sequence data meet the corresponding morphological relation includes:
if the first time sequence data comprises a first form and the second time sequence data comprises a second form corresponding to the first form within a specific time window; the position of the first aspect is determined as the transition position of the first time series data and the position of the second aspect is determined as the transition position of the second time series data.
Preferably, the determining the third time sequence data according to the first type of state data in the first time sequence data and the first type of state data in the second time sequence data includes:
fitting the first type of state data in the first time sequence data and the first type of state data in the second time sequence data to determine the third time sequence data;
the fitting process includes a data splicing process, or a data filling process.
Preferably, the first type of status data includes load status data; the fitting the first type of state data in the first time sequence data and the first type of state data in the second time sequence data to determine the third time sequence data includes:
fitting the load state data in the first time sequence data and the load state data in the second time sequence data to determine the load state data of a high-availability architecture;
load status data of the high availability architecture is determined as the third timing data.
Preferably, the first type of status data includes idle status data; the fitting the first type of state data in the first time sequence data and the first type of state data in the second time sequence data to determine the third time sequence data includes:
fitting the idle state data in the first time sequence data and the idle state data in the second time sequence data to determine the idle state data of a high-availability architecture;
and determining idle state data of the high availability architecture as the third time sequence data.
Preferably, the method further comprises:
and taking the third time sequence data as sample data, and utilizing the sample data to train the time sequence data analysis model corresponding to the high availability architecture.
Preferably, the method further comprises:
and carrying out data analysis on the third time sequence data by using a time sequence data analysis model corresponding to the high availability architecture so as to determine a corresponding analysis result.
In a second aspect, the present invention provides a processing apparatus for time series data, including:
a time sequence data determining module for determining first time sequence data of a first computing node; and second time sequence data of a second computing node matched with the first computing node;
the conversion position determining module is used for determining conversion positions in the first time sequence data and the second time sequence data when the first time sequence data and the second time sequence data meet corresponding morphological relations;
the state data determining module is used for determining first type state data in the first time sequence data and first type state data in the second time sequence data according to the conversion position;
and the time sequence data processing module is used for determining third time sequence data according to the first type of state data in the first time sequence data and the first type of state data in the second time sequence data.
In a third aspect, the present invention provides a computer-readable storage medium storing a computer program for executing the time-series data processing method of the present invention.
In a fourth aspect, the present invention provides an electronic device comprising:
a processor;
a memory for storing the processor-executable instructions;
the processor is used for reading the executable instructions from the memory and executing the instructions to realize the time sequence data processing method.
Compared with the prior art, the method and the device for processing the time sequence data, provided by the invention, have the advantages that the conversion position is determined according to the fact that the first time sequence data and the second time sequence data meet the corresponding morphological relation; determining first type state data in the first time sequence data and first type state data in the second time sequence data according to the conversion position; further determining third time sequence data; the third time sequence data can completely show the data form of the first type state in the HA architecture, so that the problem that the time sequence data in the prior art can only show the load condition of the computing node and cannot reflect the actual effective load condition of the HA architecture is avoided, and various analysis and extension application are more facilitated for the actual effective load condition of the HA architecture.
Drawings
FIG. 1 is a schematic diagram of a prior art HA architecture computation node timing data;
FIG. 2 is a flow chart of a method for processing time series data according to an embodiment of the invention;
FIG. 3 is a schematic diagram illustrating a time-series data conversion position in a time-series data processing method according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of load status data in a method for processing time series data according to an embodiment of the present invention;
fig. 5A-B are schematic diagrams of time series data fitting in a time series data processing method according to an embodiment of the present invention;
FIG. 6 is a flowchart illustrating another method for processing time-series data according to an embodiment of the present invention;
FIG. 7 is a schematic diagram of idle state data in another method for processing time-series data according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of a time-series data processing device according to an embodiment of the invention.
Detailed Description
In order to make the objects, features and advantages of the present invention more comprehensible, the technical solutions according to the embodiments of the present invention will be clearly described in the following with reference to the accompanying drawings, and it is obvious that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Typically, 2 compute nodes are included in the HA architecture, which may be referred to herein as a first compute node and a second compute node. During operation, one computing node bears the task load while the other computing node acts as a standby. If necessary, 2 nodes can be switched, i.e. the task load is quickly transferred to the standby computing node. Thereby ensuring the safety and continuity of the system operation.
Thus, each compute node in the HA architecture HAs its own timing data. The time series data is shown in fig. 1, wherein the vertical axis represents the load state of the computing node, and the horizontal axis represents time. At the location of the dashed box in fig. 1, 2 nodes of the HA architecture have switched. The load of the first computing node is reduced, and the load is changed into standby; the second computing node, in contrast, has its load raised and is turned from standby to load. That is, the time sequence data can reflect the two-dimensional relation of load/time on the computing node, intuitively embody the running condition of the HA architecture, and HAs extremely high value and diversified purposes.
The specific use of the time series data is, for example, visual presentation (i.e. as in fig. 1) for direct viewing by the staff. The method can also be used as a training sample of a time sequence data analysis model, and the trained time sequence data analysis model can predict the future load state of the HA architecture or perform data analysis of other layers. Of course, the data can also be used as input data of a time sequence data analysis model for completing training, so that the time sequence data analysis model outputs analysis results.
However, the time sequence data in fig. 1 can only represent the load condition of the computing node itself, but cannot reflect the actual effective load condition of the HA architecture, so that it is not beneficial to perform various analyses and extension applications on the actual effective load condition of the HA architecture. I.e. in the visual presentation, the staff cannot intuitively see the actual payload status of the HA architecture as a whole. In the training and analyzing process of the time sequence data analysis model, only a high load segment part in the time sequence data is usually required to be paid attention to; the low load section part may generate interference, which affects the accuracy of training and analysis of the time series data analysis model.
Therefore, an embodiment of the present invention provides a method for processing time-series data, so as to at least solve the above technical problems in the prior art. As shown in fig. 2, the method in this embodiment includes the following steps:
As described above, the first computing node and the second computing node are 2 computing nodes in the HA architecture. The corresponding first time sequence data and second time sequence data can also be shown in fig. 1. And need not be described in detail herein.
It should be noted that, in this embodiment, it is preferable that, before this step, node IP of the first computing node and the second computing node, virtual IP of the HA architecture, and node information related to an application name, an application type, and the like running on the computing nodes may be obtained from a configuration management database (i.e., CMDB, under full name Configuration Management Database) in advance. Through the above node information, it can be confirmed in advance whether the first computing node and the second computing node are matched, that is, whether they belong to the same HA architecture. If the two match, the method of this embodiment may be further performed.
As can be seen from the foregoing, when the HA architecture is switched from node to node, i.e. one computing node is turned from the load to the standby, and the other computing node is turned from the standby to the load, the corresponding time sequence data (i.e. two load curves) will necessarily exhibit a specific morphological change. This morphological change is shown in the dashed box of fig. 1 as one rise and another fall. Based on this characteristic, the transition positions in the first time series data and the second time series data will be determined in the present embodiment. The transition position may be considered as a position where the load changes drastically in the time series data, or may be considered as a time range where node switching occurs in the HA architecture.
Specifically, if the first time sequence data includes a first form and the second time sequence data includes a second form corresponding to the first form within a specific time window; the position of the first aspect is determined as the transition position of the first time series data and the position of the second aspect is determined as the transition position of the second time series data.
The first and second aspects described above refer to a form in which the timing data is suddenly increased or reduced due to node switching of the HA architecture. In practice, it may be that the first modality represents a sudden increase and the second modality represents a sudden decrease; of course, it is also possible that the first form represents a sudden decrease and the second form represents a sudden increase; the principle is the same in both cases, and it can be inferred that the HA architecture is node switched.
It should be noted that, during the HA architecture node switching process, the load transfer is completed in a shorter time. I.e., the load of one computing node drops and then the load of the other computing node rises. The entire process takes typically within seconds. That is, the time ranges of the first and second aspects representing the load change, which means HA architecture node switching, need to be within a specific time window to be considered as having a link between them. Otherwise, the load change cannot be considered to be caused by the HA architecture node switching beyond the time window; in other words, the data pattern of the load change cannot be referred to as the specific first pattern and the specific second pattern in the present embodiment.
It should be further noted that in this embodiment, a detection algorithm may be used to analyze the time series data, so as to determine whether the first configuration and the second configuration appear in the time series data. For example, it may be determined whether a numerical change between a previous data point and a subsequent data point in the time series data exceeds a certain first threshold; the first modality and the second modality may be considered to occur if a certain first threshold value is exceeded. The first and second aspects may be considered to occur if the first derivative (i.e., slope) of the load curve corresponding to the time series data is calculated and reaches a specific value. A sliding window with a specific width can be established along the transverse axis of the time sequence data, and the average value of the time sequence data in the range of the sliding window is calculated; if the average value exceeds a certain second threshold value with sliding of the sliding window, the first morphology and the second morphology may be considered to occur.
The algorithm for determining the first and second aspects may be used independently or in combination according to the requirements, and is not limited herein. In addition, the change condition of the data throughput of the first computing node and the second computing node in the time window where the first form and the second form are located can be further determined. It can be further verified in conjunction with the change in data throughput whether HA fabric node switching HAs occurred within the time window.
After the corresponding first and second patterns are determined, the position of the first pattern may be determined as the transition position of the first time series data, and the position of the second pattern may be determined as the transition position of the second time series data. As shown in fig. 3.
The transition location represents a node switch of the HA architecture. For a compute node, a transition location also means that it is changed from a load to standby, or vice versa. The time series data of the computing node can be divided into two parts based on the transition positions to represent two states (load bearing and standby) of the computing node, respectively.
The first type of state may refer to a load state of a computing node, or may refer to an idle (i.e., standby) state of the computing node, which is not limited in this regard by the present invention. In this embodiment, it may be assumed that the first type of state is a load state, that is, the first type of state data is load state data; then the idle state may be referred to as a second type of state, i.e. the second type of state data is idle state data, accordingly.
The portion of the time series data that bears the load will be first focused on in this embodiment. So that the first type of state data in the first time sequence data, namely the load state data of the first time sequence data before the transition position in fig. 3, can be obtained; and obtains the first type of state data in the second time sequence data, namely the load state data of the second time sequence data after the transition position in fig. 3. The load status data of the two parts respectively represent the running status of the respective computing nodes when bearing the load.
In this embodiment, the fitting process may be specifically performed on the first type of state data in the first time sequence data and the first type of state data in the second time sequence data to generate the third time sequence data. The third time sequence data, which is the time sequence data processed by the method in the embodiment, can completely show the data form of the first type state in the HA architecture, and exclude the second type state corresponding to the first type state.
In this embodiment, on the premise that the first type of state is a load state, the fitting process is specifically performed on load state data in the first time sequence data and load state data in the second time sequence data, so as to determine load state data actually valid for the HA architecture. The load status data actually valid by the HA architecture is the result of the timing data processing, i.e., the third timing data, as shown in fig. 4.
It should be noted that, specific modes of the fitting process may include a data splicing process, or a data filling process.
The data splicing process may be to directly splice the first type of state data in the first time sequence data and the first type of state data in the second time sequence data according to a time sequence. For example, the load state data of the first time sequence data before the transition position in fig. 3 and the load state data of the second time sequence data after the transition position are spliced, so that the load state data actually effective by the HA architecture in fig. 4 can be obtained.
However, as can be seen from fig. 4, the load curve in fig. 4 is not smooth at the splice location (as indicated by the dashed circle location), and there is some distortion. It is preferable to also perform appropriate processing on the splice position during the fitting process to make it more realistic.
If the third time series data has no periodicity characteristic in application, namely, the load value is allowed to deviate relative to the time axis, the data splicing process can be adopted. I.e. deleting part of the data in the dashed circle in fig. 4 and shifting the rest for stitching. As shown in fig. 5A. In the scenario where the third time-series data is used as the training sample, the data splicing process described above may be adopted.
If the third time series data has a periodic characteristic in application, that is, the load value is not allowed to deviate relative to the time axis, the data filling process can be adopted. Namely, deleting part of the data in the dotted circle in fig. 4, and supplementing the virtual data in the deleted part by a padding algorithm so that the load curve of the third time series data is smooth, complete and there is no offset with respect to the time axis. As shown in fig. 5B, the virtual data obtained by filling the broken line portion (straight line shape) therein. Filling dummy data according to a filling algorithm is a conventional technical means in the art, and need not be described here. In a scenario in which the third time-series data is visually presented, the above-described data filling process may be adopted.
The third time sequence data obtained in the embodiment can completely show the data form of the first type of state in the HA architecture. On the premise that the first type of state is a load state, the third time sequence data is equivalent to the load state data actually effective by the HA architecture, and the load condition of the HA architecture can be completely displayed. Therefore, the problem that the time sequence data in the prior art can only reflect the load condition of the computing node and cannot reflect the actual effective load condition of the HA architecture is avoided, and the problems of various analysis and extension application on the actual effective load condition of the HA architecture are facilitated.
According to the technical scheme, the beneficial effects of the embodiment are as follows: determining a conversion position according to the fact that the first time sequence data and the second time sequence data meet the corresponding morphological relation; determining first type state data in the first time sequence data and first type state data in the second time sequence data according to the conversion position; further determining third time sequence data; the third time sequence data can completely show the data form of the first type state in the HA architecture, so that the problem that the time sequence data in the prior art can only show the load condition of the computing node and cannot reflect the actual effective load condition of the HA architecture is avoided, and various analysis and extension application are more facilitated for the actual effective load condition of the HA architecture.
Fig. 2 shows only a basic embodiment of the method according to the invention, on the basis of which certain optimizations and developments are made, but other preferred embodiments of the method can also be obtained.
Fig. 6 shows another embodiment of the method for processing time series data according to the present invention. The present embodiment describes the method specifically on the basis of the foregoing embodiment, assuming that the first type of state is an idle state. The method specifically comprises the following steps:
The steps 601 to 602 are identical to those in the foregoing embodiments, and are not described in detail herein.
Referring to fig. 3, the present embodiment will acquire idle state data of the first timing data after the transition position and idle state data of the second timing data before the transition position in fig. 3.
In this embodiment, idle state data of the HA architecture obtained after the fitting process is shown in fig. 7. For a specific fitting process, reference is made to the specific description of step 204 above.
In this embodiment, the third timing data corresponds to idle state data of the HA architecture, and may completely represent performance of a spare part of the HA architecture.
As can be seen from the two embodiments of fig. 2 and fig. 6, the method of the present invention processes the time sequence data to obtain the load state data and the idle state data that are actually valid for the HA architecture. Unlike the prior art, which uses computing nodes as the dimension of the time sequence data, the invention uses the running state (load or idle) as the dimension of the time sequence data. By the method, time sequence data with different dimensions can be provided, so that deeper values of the time sequence data can be conveniently mined, and application and analysis of the time sequence data can be more diversified.
It should be noted that, after the third time series data is obtained in the present invention, the following applications may be performed on the third time series data, which include the following:
the third time series data as shown in fig. 4, 5A, 5B and 7 can be visually presented for the staff to browse directly. So that the staff can intuitively learn the actual effective load condition of the HA framework and the performance state of the standby part.
The third time sequence data can be used as sample data, and the sample data is utilized to train the time sequence data analysis model corresponding to the HA architecture. The time series data analysis model may include an analysis model for load state data and an analysis model for idle state data. The load state data actually valid by the HA architecture can be used as a training sample of an analytical model of the load state data; the idle state data of the HA architecture may be used as training samples for an analytical model of the idle state data. Therefore, unnecessary data is not interfered in the training process, and the training accuracy of the time sequence data analysis model can be improved.
And the data analysis can be carried out on the third time sequence data by utilizing a time sequence data analysis model corresponding to the HA architecture so as to determine a corresponding analysis result. After the training of the time sequence data analysis model is completed, the new third time sequence data can be directly used as input data of the model for the model to predict the load state or perform data analysis of other layers.
It should be noted that the HA architecture may further include a management database and a monitoring system. The configuration management database and the monitoring system also have the function of deducing the HA architecture node switching. After determining the conversion position in the time sequence data based on the method, namely, deducing the node switching of the HA architecture, the method can also mutually verify the deduced results of the configuration management database and the monitoring system, so as to more accurately judge whether the node switching of the HA architecture occurs. The verification result can be applied to the subsequent steps of the method of the present invention, and any other application necessary can be made, which is not limited in the present invention.
Fig. 8 shows an embodiment of the time series data processing device according to the present invention. The apparatus of this embodiment is a physical apparatus for performing the methods described in fig. 2 to 7. The technical solution is essentially identical to the above embodiment, and the corresponding description in the above embodiment is also applicable to this embodiment. The device in this embodiment includes:
a timing data determining module 801, configured to determine first timing data of a first computing node; and second timing data of a second computing node that matches the first computing node.
The transition position determining module 802 is configured to determine a transition position in the first time series data and the second time series data when the first time series data and the second time series data satisfy the corresponding morphological relationship.
The state data determining module 803 is configured to determine a first type of state data in the first time series data and a first type of state data in the second time series data according to the transition position.
The timing data processing module 804 is configured to determine third timing data according to the first type of state data in the first timing data and the first type of state data in the second timing data.
Additionally, based on the embodiment shown in fig. 8, it is preferable that the method further includes:
the transition position determination module 802 includes:
the form determining unit 821 is configured to determine, within a specific time window, a first form included in the first time-series data and a second form corresponding to the first form included in the second time-series data.
The position determining unit 822 is configured to determine a position of the first aspect as a transition position of the first time series data, and determine a position of the second aspect as a transition position of the second time series data.
The timing data processing module 804 includes:
the first fitting unit 841 is configured to perform fitting processing on the load status data in the first time sequence data and the load status data in the second time sequence data to determine the load status data of the high availability architecture.
A second fitting unit 842 is configured to perform fitting processing on the idle state data in the first time sequence data and the idle state data in the second time sequence data to determine the idle state data of the high availability architecture.
A third timing data determining unit 843 for determining load status data of the high availability architecture as third timing data; or, the idle state data of the high availability architecture is determined as the third timing data.
The fitting process includes a data splicing process, or a data filling process.
In addition to the methods and apparatus described above, embodiments of the invention may also be a computer program product comprising computer program instructions which, when executed by a processor, cause the processor to perform steps in a method according to various embodiments of the invention described in the "exemplary methods" section of this specification.
The computer program product may write program code for performing operations of embodiments of the present invention in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device, partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the invention may also be a computer-readable storage medium, having stored thereon computer program instructions, which when executed by a processor, cause the processor to perform steps in a method according to various embodiments of the invention described in the "exemplary method" section of the description above.
The computer readable storage medium may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium would include the following: an electrical connection having one or more wires, a portable disk, a hard disk, random Access Memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The basic principles of the present invention have been described above in connection with specific embodiments, however, it should be noted that the advantages, benefits, effects, etc. mentioned in the present invention are merely examples and not intended to be limiting, and these advantages, benefits, effects, etc. are not to be considered as essential to the various embodiments of the present invention. Furthermore, the specific details disclosed herein are for purposes of illustration and understanding only, and are not intended to be limiting, as the invention is not necessarily limited to practice with the above described specific details.
The block diagrams of the devices, apparatuses, devices, systems referred to in the present invention are only illustrative examples and are not intended to require or imply that the connections, arrangements, configurations must be made in the manner shown in the block diagrams. As will be appreciated by one of skill in the art, the devices, apparatuses, devices, systems may be connected, arranged, configured in any manner. Words such as "including," "comprising," "having," and the like are words of openness and mean "including but not limited to," and are used interchangeably therewith. The terms "or" and "as used herein refer to and are used interchangeably with the term" and/or "unless the context clearly indicates otherwise. The term "such as" as used herein refers to, and is used interchangeably with, the phrase "such as, but not limited to.
It is also noted that in the apparatus, devices and methods of the present invention, the components or steps may be disassembled and/or assembled. Such decomposition and/or recombination should be considered as equivalent aspects of the present invention.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the invention. Thus, the present invention is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit embodiments of the invention to the form disclosed herein. Although a number of example aspects and embodiments have been discussed above, a person of ordinary skill in the art will recognize certain variations, modifications, alterations, additions, and subcombinations thereof.
Claims (9)
1. A method of processing time series data, comprising:
determining first time sequence data of a first computing node; and second time sequence data of a second computing node matched with the first computing node;
if the first time sequence data comprises a first form and the second time sequence data comprises a second form corresponding to the first form within a specific time window; determining the position of the first form as a transition position of the first time series data, and determining the position of the second form as a transition position of the second time series data; the first and second configurations refer to a configuration in which the time-series data is suddenly increased or suddenly decreased due to node switching of the high-availability HA architecture;
determining first-type state data in the first time sequence data and first-type state data in the second time sequence data according to the conversion position;
fitting the first type of state data in the first time sequence data and the first type of state data in the second time sequence data to determine third time sequence data;
the first type of status data includes load status data or idle status data.
2. The method of claim 1, the fitting process comprising a data stitching process, a data splicing process, or a data padding process.
3. The method of claim 1, the first type of status data comprising load status data; the fitting the first type of state data in the first time sequence data and the first type of state data in the second time sequence data to determine the third time sequence data includes:
fitting the load state data in the first time sequence data and the load state data in the second time sequence data to determine the load state data of a high-availability architecture;
load status data of the high availability architecture is determined as the third timing data.
4. The method of claim 1, the first type of status data comprising idle status data; the fitting the first type of state data in the first time sequence data and the first type of state data in the second time sequence data to determine the third time sequence data includes:
fitting the idle state data in the first time sequence data and the idle state data in the second time sequence data to determine the idle state data of a high-availability architecture;
and determining idle state data of the high availability architecture as the third time sequence data.
5. The method of any one of claims 1-4, further comprising:
and taking the third time sequence data as sample data, and utilizing the sample data to train the time sequence data analysis model corresponding to the high availability architecture.
6. The method of any one of claims 1-4, further comprising:
and carrying out data analysis on the third time sequence data by using a time sequence data analysis model corresponding to the high availability architecture so as to determine a corresponding analysis result.
7. A processing apparatus of time series data, comprising:
a time sequence data determining module for determining first time sequence data of a first computing node; and second time sequence data of a second computing node matched with the first computing node;
the conversion position determining module is used for including a first form in the first time sequence data and including a second form corresponding to the first form in the second time sequence data in a specific time window; determining the position of the first form as a transition position of the first time series data, and determining the position of the second form as a transition position of the second time series data; the first and second configurations refer to a configuration in which the time-series data is suddenly increased or suddenly decreased due to node switching of the high-availability HA architecture;
the state data determining module is used for determining first type state data in the first time sequence data and first type state data in the second time sequence data according to the conversion position;
the time sequence data processing module is used for carrying out fitting processing on the first type of state data in the first time sequence data and the first type of state data in the second time sequence data so as to determine third time sequence data;
the first type of state data includes load type state data or idle type state data.
8. A computer-readable storage medium storing a computer program for executing the method of processing time-series data according to any one of the preceding claims 1-6.
9. An electronic device, comprising:
a processor;
a memory for storing the processor-executable instructions;
the processor is configured to read the executable instructions from the memory and execute the instructions to implement the method of processing time series data according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011539336.7A CN112685246B (en) | 2020-12-23 | 2020-12-23 | Time sequence data processing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011539336.7A CN112685246B (en) | 2020-12-23 | 2020-12-23 | Time sequence data processing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112685246A CN112685246A (en) | 2021-04-20 |
| CN112685246B true CN112685246B (en) | 2023-04-28 |
Family
ID=75451179
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011539336.7A Active CN112685246B (en) | 2020-12-23 | 2020-12-23 | Time sequence data processing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112685246B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016123972A1 (en) * | 2015-02-02 | 2016-08-11 | 中兴通讯股份有限公司 | Load balancing method and load balancing apparatus |
| CN109408254A (en) * | 2018-09-30 | 2019-03-01 | 上海与德科技有限公司 | A kind of information processing method, system and server |
| CN109936473A (en) * | 2017-12-19 | 2019-06-25 | 华耀(中国)科技有限公司 | Distributed computing system and its operation method based on deep learning prediction |
| CN110149237A (en) * | 2019-06-13 | 2019-08-20 | 东北大学 | A kind of Hadoop platform calculate node load predicting method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200034745A1 (en) * | 2015-10-19 | 2020-01-30 | Nutanix, Inc. | Time series analysis and forecasting using a distributed tournament selection process |
-
2020
- 2020-12-23 CN CN202011539336.7A patent/CN112685246B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016123972A1 (en) * | 2015-02-02 | 2016-08-11 | 中兴通讯股份有限公司 | Load balancing method and load balancing apparatus |
| CN109936473A (en) * | 2017-12-19 | 2019-06-25 | 华耀(中国)科技有限公司 | Distributed computing system and its operation method based on deep learning prediction |
| CN109408254A (en) * | 2018-09-30 | 2019-03-01 | 上海与德科技有限公司 | A kind of information processing method, system and server |
| CN110149237A (en) * | 2019-06-13 | 2019-08-20 | 东北大学 | A kind of Hadoop platform calculate node load predicting method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112685246A (en) | 2021-04-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210136098A1 (en) | Root cause analysis in multivariate unsupervised anomaly detection | |
| US10599367B2 (en) | Updating storage migration rates | |
| CN110516971A (en) | Method, apparatus, medium and the calculating equipment of abnormality detection | |
| CN111552633A (en) | Interface abnormal call testing method and device, computer equipment and storage medium | |
| CN109460396B (en) | Model processing method and device, storage medium and electronic equipment | |
| CN114580263A (en) | Knowledge graph-based information system fault prediction method and related equipment | |
| CN110221145A (en) | Fault Diagnosis for Electrical Equipment method, apparatus and terminal device | |
| US20170212928A1 (en) | Cognitive decision making based on dynamic model composition | |
| US11620486B2 (en) | Estimating and visualizing collaboration to facilitate automated plan generation | |
| JP2021128779A (en) | Method, device, apparatus, and storage medium for expanding data | |
| CN116701031A (en) | Root cause model training method, analysis method and device in micro-service system | |
| CN109643087B (en) | System and method for selecting controllable parameters for operational safety of a device | |
| US20230133541A1 (en) | Alert correlating using sequence model with topology reinforcement systems and methods | |
| CN112801434A (en) | Method, device, equipment and storage medium for monitoring performance index health degree | |
| CN113419941A (en) | Evaluation method and apparatus, electronic device, and computer-readable storage medium | |
| CN112685246B (en) | Time sequence data processing method and device | |
| CN110782117A (en) | Driving risk detection method and device, storage medium and electronic equipment | |
| CN114141236B (en) | Language model updating method and device, electronic equipment and storage medium | |
| CN111815442B (en) | Link prediction method and device and electronic equipment | |
| US20230061222A1 (en) | Early stopping of artificial intelligence model training using control limits | |
| Sagaama et al. | Automatic parameter tuning for big data pipelines with deep reinforcement learning | |
| CN113190352B (en) | General CPU-oriented deep learning calculation acceleration method and system | |
| CN111679924B (en) | Reliability simulation method and device for componentized software system and electronic equipment | |
| CN114996136A (en) | Test data determination method and device, electronic equipment and readable storage medium | |
| CN112733015A (en) | User behavior analysis method, device, equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |