A Distributed-Based Multimodal High-Efficiency Temporal Data Storage Method

Abstract

The present invention relates to the technical field of time-series data storage, and discloses a distributed multi-modal efficient time-series data storage method. By collecting the amount of data generated per unit time of each time-series data generation device, analyzing the real-time data storage amount and the remaining service life value of each distributed time-series data storage device, determining the target update storage device and the non-target update storage device, performing the generation of the storage path from the target update storage device and the non-target update storage device to the time-series data generation device, and controlling the time-series data storage action of the time-series data generation device. Thus, by considering multiple influencing factors of time-series data and storage devices, it can drive the life management and data migration of several storage devices to fit the update cycle of each storage device, reduce the difficulty of the distributed data storage system in terms of time-series data storage allocation and storage device management, and improve the data storage capacity and stability of the time-series database.

Description

Technical Field

[0001] The present invention relates to the technical field of time-series data storage, and particularly to a distributed multi-modal efficient time-series data storage method. Background Art

[0002] Time-series databases are widely used in fields such as the Internet of Things, finance, and industrial control, mainly for storing and querying large-scale time-series data. In the prior art, time-series database systems have efficient data insertion and query capabilities, but there are challenges in data storage and scalability. Distributed storage systems provide reliable storage solutions, but there are still many limitations in the time-series data storage allocation and storage device management of distributed storage systems.

[0003] Specifically, in a distributed data storage system, different distributed time-series data storage devices may have different storage capacities (usually referring to the number of time-series data groups of unit size that can be stored in a storage device) and storage device service lives (usually referring to the maximum number of write operations for a time-series data group of unit size in a storage device). Different data source ends select different distributed time-series data storage devices to store data in the distributed storage system, and the amount of data generated per unit time is also different. This makes the life management accuracy of several distributed time-series data storage devices inside the distributed data storage system not high, and it is easy to cause the service lives of multiple distributed time-series data storage devices to reach the upper limit at the same time, requiring simultaneous replacement of storage devices and data migration, rather than reaching the upper limit sequentially and performing storage device replacement and data migration sequentially. Such a situation will affect the data storage capacity of the entire time-series database, and further affect the business operation of the actual application scenario. At the same time, due to different data retention requirements of different data source ends, the expiration deletion times of the time-series data generated by each time-series data generation device when stored in the distributed time-series data storage device are different, which further increases the difficulty of the distributed data storage system in time-series data storage allocation and storage device management.

[0004] Therefore, how to reduce the difficulty of the distributed data storage system in time-series data storage allocation and storage device management, improve the data storage capacity and stability of the time-series database, and realize the high-precision life management of storage devices and the efficient process of device replacement and data migration in the distributed data storage system is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] The present invention provides a distributed multi-modal efficient time-series data storage method, aiming to solve at least one of the above technical problems.

[0006] To achieve the above object, the present invention provides a distributed multi-modal efficient time-series data storage method, and the method includes the following steps:

[0007] S1: Collect the time-series data characteristics of each time-series data generation device within the collection area, and generate the data generation amount per unit time of each time-series data generation device according to the time-series data characteristics.

[0008] S2: Obtain the device identifiers and storage information of each distributed time-series data storage device in the distributed data storage system, and determine the remaining service life value of the storage device and the real-time data storage amount of several groups of time-series data for each distributed time-series data storage device.

[0009] S3: Determine the target updated storage devices and non-target updated storage devices for the current storage device update cycle according to the remaining service life value of the storage device of each distributed time-series data storage device.

[0010] S4: Generate a time-series data storage matching strategy for each distributed time-series data storage device based on the remaining service life value of the storage device, the real-time data storage amount of the target updated storage device and the non-target updated storage device, and the data generation amount per unit time of each time-series data generation device; wherein, the time-series data storage matching strategy includes the first time-series data storage paths of several first time-series data generation devices and each target updated storage device and the second time-series data storage paths of several second time-series data generation devices and each non-target updated storage device.

[0011] S5: Control each time-series data generation device to perform corresponding time-series data storage actions within the current storage device update cycle based on the time-series data storage matching strategy.

[0012] Optionally, step S1 specifically includes:

[0013] S11: Collect the time-series data characteristics of each time-series data generation device within the collection area; wherein, the time-series data characteristics include the time-series data type and the time-series data upload frequency.

[0014] S12: Determine the data volume size of each time-series data generation device to perform a single time-series data transmission according to the time-series data type, and calculate the data generation amount per unit time of each time-series data generation device by using the data volume size and the time-series data upload frequency.

[0015] Optionally, step S2 specifically includes:

[0016] S21: Obtain the device identifiers and storage information of each distributed time-series data storage device in the distributed data storage system; wherein, the storage information includes the historical storage data volume and the real-time storage information.

[0017] S22: Calculate the remaining service life value of each distributed time-series data storage device by using the historical storage data volume in the device identifier and storage information;

[0018] S23: Determine the real-time data storage volume of each distributed time-series data storage device by using the real-time storage information in the storage information.

[0019] Optionally, step S22 specifically includes:

[0020] S221: Query the initial service life value of the storage device of each distributed time-series data storage device by using the device identifier, and extract the historical storage data volume and real-time data storage information in the storage information;

[0021] S222: Calculate the remaining service life value of the storage device of each distributed time-series data storage device according to the corresponding relationship between the historical storage data volume and the reduction value of the service life of the storage device by using the historical storage data volume in the storage information and the initial service life value of the storage device.

[0022] Optionally, step S23 specifically includes:

[0023] S231: Determine the time-series data types of several groups of time-series data stored in each distributed time-series data storage device by using the real-time data storage information;

[0024] S232: Calculate the real-time data storage volume in each distributed time-series data storage device according to the time-series data type of each group of time-series data.

[0025] Optionally, step S3 specifically includes:

[0026] S31: Sort the remaining service life values of the storage devices of each distributed time-series data storage device from low to high to generate a sorted list of the remaining life values of the distributed time-series data storage devices;

[0027] S32: Use the first distributed time-series data storage device in the sorted list of the remaining life values as the target updated storage device for the current storage device update cycle, and use the remaining distributed time-series data storage devices as non-target updated storage devices.

[0028] Optionally, step S4 specifically includes:

[0029] S41: Based on the remaining service life value of the target updated storage device and the duration of the current storage device update cycle, calculate the first unit time data storage volume when the target updated storage device receives time-series data in the current storage device on the principle that the remaining service life value of the target updated storage device at the end of the current storage device update cycle is lower than the device update threshold;

[0030] S42: Based on the first unit time data storage amount and the unit time data generation amount of each time series data generation device, match several first time series data generation devices among all time series data generation devices whose sum of unit time data generation amounts is closest to the first unit time data storage amount;

[0031] S43: Add the first time series data storage paths of the several first time series data generation devices and the target update storage device to the time series data storage matching policy as the time series data storage methods of the several first time series data generation devices;

[0032] S44: Based on the first ratio of the remaining service life values of the non-target update storage devices, with the principle that the second ratio of the sum of the unit time data generation amounts assigned to each non-target update storage device by the remaining time series data generation devices is closest to the first ratio, match several second time series data generation devices for each non-target update storage device among the remaining time series data generation devices;

[0033] S45: Add the second time series data storage paths of the several second time series data generation devices and the corresponding non-target update storage devices to the time series data storage matching policy as the time series data storage methods of the several second time series data generation devices.

[0034] Optionally, step S4 further includes:

[0035] S46: Obtain the data storage period of each time series data generation device and the data storage amount upper limit value of each distributed time series data storage device and the data timestamp information of each group of stored time series data;

[0036] S47: According to the data storage period and the data timestamp information, considering the real-time data storage amount and the data storage amount upper limit value of each distributed time series data storage device, determine whether to re-execute the determination of the target update storage device and the non-target update storage device.

[0037] Optionally, step S47 specifically includes:

[0038] S471: According to the data storage period and the data timestamp information, calculate the regular deletion time of each group of time series data in each distributed time series data storage device, and considering the real-time data storage amount and the data storage amount upper limit value of each distributed time series data storage device, determine whether the real-time data storage amount will be higher than the data storage amount upper limit value when each distributed time series data storage device executes the time series data storage of the time series data generation device according to the corresponding time series data storage path in the time series data storage matching policy;

[0039] S472: If so, remove the distributed time-series data storage device, and return to step S3 to re-determine the target updated storage device and the non-target updated storage device for execution.

[0040] Optionally, the method further includes step S6: At the end of the current storage device update cycle, import the data in the target updated storage device into the newly connected distributed time-series data storage device, and re-determine the time-series data storage matching policy for the next storage device update cycle.

[0041] The beneficial effects of the present invention are as follows: A distributed multi-modal efficient time-series data storage method is proposed. By considering multiple influencing factors from time-series data to storage devices, first generate the storage path of the target updated storage device based on the current storage device update cycle, and then generate the storage path of the non-target updated storage device based on the real-time data storage volume ratio, so as to drive the life management and data migration of several storage devices in the distributed data storage system to fit each storage device update cycle, reduce the difficulty of the distributed data storage system in time-series data storage allocation and storage device management, improve the data storage capacity and stability of the time-series database, and realize the high-precision life management and efficient device replacement and data migration process of storage devices in the distributed data storage system. Description of the Drawings

[0042] Figure 1 It is a flowchart of an embodiment of the distributed multi-modal efficient time-series data storage method of the present invention.

[0043] The realization, functional characteristics and advantages of the object of the present invention will be further described in conjunction with the embodiments with reference to the drawings. Detailed Embodiments

[0044] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0045] The embodiment of the present invention provides a distributed multi-modal efficient time-series data storage method, referring to Figure 1 , Figure 1 It is a flowchart of an embodiment of the distributed multi-modal efficient time-series data storage method of the present invention.

[0046] In this embodiment, a distributed multi-modal efficient time-series data storage method, the method includes the following steps:

[0047] S1: Collect the time-series data characteristics of each time-series data generation device within the collection area, and generate the amount of data generated per unit time for each time-series data generation device according to the time-series data characteristics;

[0048] S2: Obtain the device identifiers and storage information of each distributed time-series data storage device in the distributed data storage system, and determine the remaining service life value of the storage device and the real-time data storage amount of several groups of time-series data for each distributed time-series data storage device;

[0049] S3: Determine the target updated storage devices and non-target updated storage devices for the current storage device update cycle according to the remaining service life value of the storage device of each distributed time-series data storage device;

[0050] S4: Generate a time-series data storage matching strategy for each distributed time-series data storage device based on the remaining service life value of the storage device, the real-time data storage amount of the target updated storage device and the non-target updated storage device, and the amount of data generated per unit time for each time-series data generation device; wherein, the time-series data storage matching strategy includes the first time-series data storage paths of several first time-series data generation devices and each target updated storage device and the second time-series data storage paths of several second time-series data generation devices and each non-target updated storage device;

[0051] S5: Control each time-series data generation device to perform corresponding time-series data storage actions within the current storage device update cycle based on the time-series data storage matching strategy.

[0052] It should be noted that since different distributed time-series data storage devices in the distributed data storage system may have different storage capacities (usually referring to the storage quantity of time-series data groups of unit size in the storage device), storage device service lives (usually referring to the maximum number of write operations of time-series data groups of unit size in the storage device), different distributed time-series data storage devices are selected for storing distributed time-series data from different data source ends, and the amount of data generated per unit time is also different, the accuracy of life management of several distributed time-series data storage devices inside the distributed data storage system is not high, which easily causes the service lives of multiple distributed time-series data storage devices to reach the upper limit at the same time and requires simultaneous replacement of storage devices and data migration, rather than reaching the upper limit in sequence and performing replacement of storage devices and data migration in sequence. Such a situation will affect the data storage capacity of the entire time-series database, and further affect the business operation of the actual application scenario. At the same time, due to different data retention requirements of different data source ends, the expiration deletion times are different when the time-series data generated by each time-series data generation device is stored in the distributed time-series data storage device, which further increases the difficulty of time-series data storage allocation and storage device management in the distributed data storage system.

[0053] To solve the above problems, in this embodiment, the data generation amount per unit time of each time-series data generation device is collected, the real-time data storage amount and the remaining service life value of the storage device of each distributed time-series data storage device are analyzed, the target update storage device and the non-target update storage device are determined, the storage path generation from the target update storage device and the non-target update storage device to the time-series data generation device is executed, and the time-series data storage action of the time-series data generation device is controlled. Thus, by considering multiple influencing factors of time-series data and storage devices, the life management and data migration of several storage devices in the distributed data storage system can fit the update cycle of each storage device, the difficulty of the distributed data storage system in terms of time-series data storage allocation and storage device management can be reduced, and the data storage capacity and stability of the time-series database can be improved.

[0054] In a preferred embodiment, step S1 specifically includes:

[0055] S11: Collect the time-series data characteristics of each time-series data generation device within the regional scope; wherein, the time-series data characteristics include the time-series data type and the time-series data upload frequency;

[0056] S12: According to the time-series data type, determine the data volume size of each time-series data generation device for performing a single time-series data transmission, and use the data volume size and the time-series data upload frequency to calculate the data generation amount per unit time of each time-series data generation device.

[0057] In this embodiment, by collecting the time-series data type and the time-series data upload frequency when each time-series data generation device uploads time-series data within the regional scope, according to the time-series data type, the data volume size of a single time-series data transmission is determined, and then the data generation amount per unit time of each time-series data generation device is calculated by multiplying the time-series data upload frequency by the data volume size of a single time-series data transmission. By accurately analyzing the data generation amount per unit time of each time-series data generation device, data support is provided for the subsequent high-precision management of the storage device life and the generation of the data storage allocation strategy.

[0058] In a preferred embodiment, step S2 specifically includes:

[0059] S21: Obtain the device identifier and storage information of each distributed time-series data storage device in the distributed data storage system; wherein, the storage information includes the historical storage data volume and the real-time storage information;

[0060] S22: Use the device identifier and the historical storage data volume in the storage information to calculate the remaining service life value of each distributed time-series data storage device;

[0061] S23: Determine the real-time data storage capacity of each distributed time-series data storage device by using the real-time storage information in the stored information.

[0062] Furthermore, step S22 specifically includes:

[0063] S221: Query the initial service life value of each distributed time-series data storage device by using the device identifier, and extract the historical storage data volume and real-time data storage information in the stored information;

[0064] S222: Calculate the remaining service life value of each distributed time-series data storage device according to the corresponding relationship between the historical storage data volume and the reduction value of the service life of the storage device by using the historical storage data volume in the stored information and the initial service life value of the storage device.

[0065] Furthermore, step S23 specifically includes:

[0066] S231: Determine the time-series data types of several groups of time-series data stored in each distributed time-series data storage device by using the real-time data storage information;

[0067] S232: Calculate the real-time data storage capacity in each distributed time-series data storage device according to the time-series data types of each group of time-series data.

[0068] In this embodiment, first obtain the device identifier and stored information of each distributed time-series data storage device, query the initial service life value of the storage device by using the device identifier, and then calculate the remaining service life value of the storage device by converting the historical storage data volume into the reduction value of the service life of the storage device; then, determine the time-series data type according to the real-time data storage information, and calculate the real-time data storage capacity in the distributed time-series data storage device.

[0069] In a preferred embodiment, step S3 specifically includes:

[0070] S31: Sort the remaining service life values of each distributed time-series data storage device from low to high to generate a sorted list of the remaining life values of the distributed time-series data storage devices;

[0071] S32: Use the first distributed time-series data storage device in the sorted list of the remaining life values as the target update storage device for the current storage device update cycle, and use the remaining distributed time-series data storage devices as non-target update storage devices.

[0072] In this embodiment, the storage device with the lowest remaining life value is used as the target update storage device, and the remaining storage devices are used as non-target update storage devices. It should be noted that the target update storage device is the storage device that needs to be replaced and have its data migrated after the end of the current storage update cycle, while the non-target update storage devices are the storage devices for which replacement and data migration will be carried out in subsequent storage update cycles.

[0073] In a preferred embodiment, step S4 specifically includes:

[0074] S41: Based on the remaining service life value of the target update storage device and the duration of the current storage device update cycle, and taking the principle that the remaining service life value of the target update storage device at the end of the current storage device update cycle is lower than the device update threshold, calculate the first unit time data storage amount of the target update storage device when receiving time-series data in the current storage device.

[0075] S42: According to the first unit time data storage amount and the unit time data generation amount of each time-series data generating device, match several first time-series data generating devices whose sum of the unit time data generation amounts is closest to the first unit time data storage amount among all time-series data generating devices.

[0076] S43: Add the several first time-series data generating devices and the first time-series data storage paths of the target update storage device to the time-series data storage matching policy as the time-series data storage methods of the several first time-series data generating devices.

[0077] S44: Based on the first ratio of the remaining service life value of the non-target update storage device, and taking the principle that the second ratio of the sum of the unit time data generation amounts assigned to each non-target update storage device by the remaining time-series data generating devices is closest to the first ratio, match several second time-series data generating devices for each non-target update storage device among the remaining time-series data generating devices.

[0078] S45: Add the several second time-series data generating devices and the second time-series data storage paths of the corresponding non-target update storage devices to the time-series data storage matching policy as the time-series data storage methods of the several second time-series data generating devices.

[0079] In this embodiment, by considering multiple influencing factors of time-series data to the storage device, first execute the generation of the storage path for the target updated storage device based on the current storage device update cycle, and then execute the generation of the storage path for the non-target updated storage device based on the real-time data storage volume ratio, so as to drive the life management and data migration of several storage devices in the distributed data storage system to fit each storage device update cycle, reduce the difficulty of the distributed data storage system in terms of time-series data storage allocation and storage device management, improve the data storage capacity and stability of the time-series database, and realize the process of high-precision life management of storage devices and efficient device replacement and data migration in the distributed data storage system.

[0080] In a preferred embodiment, step S4 further includes:

[0081] S46: Obtain the data storage period of each time-series data generation device, the data storage volume upper limit value of each distributed time-series data storage device, and the data timestamp information of each group of time-series data stored.

[0082] S47: According to the data storage period and data timestamp information, considering the real-time data storage volume and the data storage volume upper limit value of each distributed time-series data storage device, determine whether to re-execute the determination of the target updated storage device and the non-target updated storage device.

[0083] Furthermore, step S47 specifically includes:

[0084] S471: According to the data storage period and data timestamp information, calculate the regular deletion time of each group of time-series data in each distributed time-series data storage device. Considering the real-time data storage volume and the data storage volume upper limit value of each distributed time-series data storage device, determine whether the real-time data storage volume will exceed the data storage volume upper limit value when each distributed time-series data storage device executes the time-series data storage of the time-series data generation device according to the time-series data storage path corresponding to the time-series data storage matching strategy.

[0085] S472: If so, exclude this distributed time-series data storage device, and return to step S3 to re-execute the determination of the target updated storage device and the non-target updated storage device.

[0086] Considering that the data retention requirements of different data source ends are different, the expiration deletion times of the time-series data generated by each time-series data generation device when storing the time-series data in the distributed time-series data storage device are different, which further increases the difficulty of the distributed data storage system in terms of time-series data storage allocation and storage device management. In this embodiment, after generating the time-series data storage matching policy according to the above steps, it is necessary to determine whether, when each distributed time-series data storage device executes time-series data storage according to this time-series data storage matching policy, the real-time data storage volume of a certain distributed time-series data storage device is higher than the upper limit value of the data storage volume (since the time-series data is continuously imported and stored, and at the same time, the expiration deletion is also continuously carried out). If so, it indicates that the current time-series data storage matching policy cannot meet the space limit of the storage device. At this time, it is necessary to temporarily exclude this distributed time-series data storage device from the current storage update cycle (because there is a large amount of new data in this distributed time-series data storage device, and the balance of the data storage volume cannot be maintained by the way of expiration deletion and importing time-series data at the same time). After exclusion, return to step S3 to execute the determination of the target update storage device and the non-target update storage device and the subsequent generation of the time-series data storage matching policy.

[0087] In a preferred embodiment, the method further includes step S6: at the end of the current storage device update cycle, import the data in the target update storage device into the newly accessed distributed time-series data storage device, and re-determine the time-series data storage matching policy for the next storage device update cycle.

[0088] In this embodiment, after obtaining the determined time-series data storage matching policy, execute time-series data storage according to this time-series data storage matching policy. After that, at the end of the current storage device update cycle, perform the replacement of the storage device and the migration of the stored data, and re-determine the new time-series data storage matching policy, so as to realize the storage allocation of the time-series data and the storage device management in the distributed data storage system.

[0089] In addition, the present invention also proposes a distributed multi-modal efficient time-series data storage device, and the distributed multi-modal efficient time-series data storage device includes: a memory, a processor, and a distributed multi-modal efficient time-series data storage program stored on the memory and executable on the processor. When the distributed multi-modal efficient time-series data storage program is executed by the processor, the steps of the above-mentioned distributed multi-modal efficient time-series data storage method are implemented.

[0090] The specific implementation manners of the distributed multi-modal efficient time-series data storage device of the present application are basically the same as those of the above-mentioned embodiments of the distributed multi-modal efficient time-series data storage method, and will not be elaborated herein.

[0091] It should be understood that in the description of this specification, the descriptions with reference to terms such as "one embodiment", "another embodiment", "other embodiments", or "the first embodiment to the Nth embodiment" mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0092] It should be noted that in this article, the term "comprising", "including", or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article, or system including a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article, or system. Without further limitation, an element defined by the statement "including an..." does not exclude the existence of additional identical elements in the process, method, article, or system including that element.

[0093] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structural or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A distributed multi-modal efficient time series data storage method, characterized in that: The method comprises the following steps: S1: collecting time series data features of each time series data generating device within the area, and generating a unit time data generation amount of each time series data generating device according to the time series data features; S2: Obtaining the device identification and storage information of each distributed time series data storage device in the distributed data storage system, and determining the remaining service life value of the storage device of each distributed time series data storage device and the real-time data storage capacity of several groups of time series data; S3: determining the target update storage device and the non-target update storage device in the current storage device update cycle according to the remaining service life value of the storage device of each distributed time-series data storage device; S4: Based on the remaining service life values ​​of the target update storage device and the non-target update storage device, the real-time data storage volume, and the unit time data generation volume of each time series data generation device, a time series data storage matching strategy for each distributed time series data storage device is generated; wherein the time series data storage matching strategy includes a first time series data storage path between a plurality of first time series data generation devices and each target update storage device and a second time series data storage path between a plurality of second time series data generation devices and each non-target update storage device; specifically includes: S41: Based on the remaining service life value of the target updated storage device and the duration of the current storage device update cycle, and on the principle that the remaining service life value of the target updated storage device at the end of the current storage device update cycle is lower than the device update threshold, calculate the first unit time data storage amount when the target updated storage device receives the time series data in the current storage device; S42: According to the first unit time data storage capacity and the unit time data generation capacity of each time series data generation device, matching a plurality of first time series data generation devices whose sum of unit time data generation capacity is closest to the first unit time data storage capacity among all time series data generation devices; S43: adding the first time series data storage paths of the plurality of first time series data generating devices and the target update storage device to the time series data storage matching strategy as the time series data storage mode of the plurality of first time series data generating devices; S44: Based on the first ratio of the remaining service life values ​​of the non-target update storage devices, a plurality of second time series data generating devices are matched for each non-target update storage device in the remaining time series data generating devices, on the principle that the second ratio of the sum of the unit time data generation amount allocated by the remaining time series data generating devices to each non-target update storage device is closest to the first ratio; S45: adding the second time series data storage paths of the plurality of second time series data generating devices and the corresponding non-target update storage devices to the time series data storage matching strategy as a time series data storage mode for the plurality of second time series data generating devices; S5: Based on the time series data storage matching strategy, control each time series data generating device to execute a corresponding time series data storage action within the current storage device update cycle.

2. The distributed multi-modal efficient time series data storage method according to claim 1, characterized in that: Step S1 specifically includes: S11: collecting time series data characteristics of each time series data generating device within the area; wherein the time series data characteristics include the time series data type and the time series data upload frequency; S12: Determine the data volume of a single time series data transmission executed by each time series data generating device according to the time series data type, and calculate the data generation volume per unit time of each time series data generating device by using the data volume and the time series data upload frequency.

3. The distributed multi-modal efficient time series data storage method according to claim 1, characterized in that: Step S2 specifically includes: S21: Obtaining a device identification and storage information of each distributed time series data storage device in the distributed data storage system; wherein the storage information includes a historical storage data volume and real-time storage information; S22: Calculate the remaining service life value of the storage device of each distributed time series data storage device by using the device identification and the amount of historical storage data in the storage information; S23: Determine the real-time data storage capacity of each distributed time series data storage device by using the real-time storage information in the storage information.

4. The distributed multi-modal efficient time series data storage method according to claim 3, characterized in that: Step S22 specifically includes: S221: using the device identifier, querying the initial service life value of the storage device of each distributed time series data storage device, and extracting the historical storage data volume and real-time data storage information in the storage information; S222: Calculate the remaining service life of each distributed time series data storage device according to the corresponding relationship between the amount of historical storage data and the reduced service life of the storage device by using the amount of historical storage data in the storage information and the initial service life of the storage device.

5. The distributed multi-modal efficient time series data storage method according to claim 4, characterized in that: Step S23 specifically includes: S231: using the real-time data storage information, determining the time series data types of several groups of time series data stored in each distributed time series data storage device; S232: Calculate the real-time data storage capacity in each distributed time series data storage device according to the time series data type of each group of time series data.

6. The distributed multi-modal efficient time series data storage method according to claim 1, characterized in that: Step S3 specifically includes: S31: sorting the remaining service life value of each distributed time series data storage device from low to high to generate a sorted list of remaining service life values ​​of the distributed time series data storage devices; S32: The first distributed time series data storage device in the remaining life value sorted list is used as a target update storage device in the current storage device update cycle, and the remaining distributed time series data storage devices are used as non-target update storage devices.

7. The distributed multi-modal efficient time series data storage method according to claim 6, characterized in that: Step S4 specifically includes: S41: Based on the remaining service life value of the target updated storage device and the duration of the current storage device update cycle, and on the principle that the remaining service life value of the target updated storage device at the end of the current storage device update cycle is lower than the device update threshold, calculate the first unit time data storage amount when the target updated storage device receives the time series data in the current storage device; S42: According to the first unit time data storage capacity and the unit time data generation capacity of each time series data generation device, matching a plurality of first time series data generation devices whose sum of unit time data generation capacity is closest to the first unit time data storage capacity among all time series data generation devices; S43: adding the first time series data storage paths of the plurality of first time series data generating devices and the target update storage device to the time series data storage matching strategy as the time series data storage mode of the plurality of first time series data generating devices; S44: Based on the first ratio of the remaining service life values ​​of the non-target update storage devices, a plurality of second time series data generating devices are matched for each non-target update storage device in the remaining time series data generating devices, on the principle that the second ratio of the sum of the unit time data generation amount allocated by the remaining time series data generating devices to each non-target update storage device is closest to the first ratio; S45: adding the second time series data storage paths of the plurality of second time series data generating devices and the corresponding non-target update storage devices to the time series data storage matching strategy as the time series data storage mode of the plurality of second time series data generating devices.

8. The distributed multi-modal efficient time series data storage method according to claim 7, characterized in that: Step S4 also includes: S46: Obtaining the data storage period of each time series data generating device, the upper limit of the data storage capacity of each distributed time series data storage device, and the data stamp information of each set of stored time series data; S47: Based on the data storage period and data stamp information, and taking into account the real-time data storage capacity and the upper limit of the data storage capacity of each distributed time series data storage device, it is determined whether to re-execute the determination of the target update storage device and the non-target update storage device.

9. The distributed multi-modal efficient time series data storage method according to claim 8, characterized in that: The step S47 specifically includes: S471: Calculate the periodic deletion time of each group of time series data in each distributed time series data storage device according to the data storage period and data stamp information, consider the real-time data storage capacity and the data storage capacity upper limit of each distributed time series data storage device, and determine whether the real-time data storage capacity of each distributed time series data storage device is higher than the data storage capacity upper limit when the time series data storage path corresponding to the time series data storage matching strategy is executed by the time series data storage device of the time series data generation device; S472: If yes, remove the distributed time series data storage device, and return to step S3 to determine the target update storage device and the non-target update storage device.

10. The distributed multi-modal efficient time series data storage method according to claim 1, characterized in that: The method further includes step S6: when the current storage device update cycle ends, the data in the target update storage device is imported into the newly connected distributed time series data storage device, and the time series data storage matching strategy for the next storage device update cycle is re-determined.

Images