CN111372095B

CN111372095B - Method and device for calculating heat

Info

Publication number: CN111372095B
Application number: CN201811593023.2A
Authority: CN
Inventors: 徐佳宏; 陈华兵
Original assignee: Shenzhen Ipanel TV Inc
Current assignee: Shenzhen Ipanel TV Inc
Priority date: 2018-12-25
Filing date: 2018-12-25
Publication date: 2023-06-23
Anticipated expiration: 2038-12-25
Also published as: CN111372095A

Abstract

The application provides a method and a device for calculating heat, comprising the following steps: monitoring the number of times that a target file is requested to be accessed by a terminal and is not cached by an edge server in a preset heat calculation period; the edge server is a cache server between the terminal and the resource server; and if the times reach a threshold value, calculating the heat value of the target file at the current moment based on the heat calculation period. According to the method and the device, the heat calculation period is set for heat calculation, so that the calculated heat value is closer to the actual heat at the current moment, and the accuracy of heat calculation is improved.

Description

Method and device for calculating heat

Technical Field

The invention relates to the technical field of Internet, in particular to a method and a device for calculating heat.

Background

With the increasing development of the internet, more and more films, data, files and the like can be easily found on the internet, so that people can find things needed by themselves without going home. As people use the internet more and more, so too is the demand for it. When a user uses a streaming media resource, the faster and better the access speed is required. In the prior art, the cache acceleration is usually performed by using a CDN (Content DeliveryNetwork ), and the access acceleration is performed by storing a part of resources from a source station into a CDN node in advance and shortening the distance between the source station and a user. In the existing CDN cache acceleration scheme, when caching resource files to CDN nodes, resources with high heat are usually selected and stored.

However, in the existing heat calculation method, the total access amount divided by the time is calculated, so if the historical playing access number of a certain resource file is higher or lower, even if the playing frequency is continuously increased and reduced along with the time, the heat value calculated by the scheme is not changed greatly due to the early influence effect of the accumulated value, that is, still in a high heat state or in a low heat state, so that the file becomes a cold film (no on demand during a period of time) but is not replaced by a CDN node, or becomes a hot film (the on demand amount during a period of time is higher) but cannot be buffered to the CDN node.

Therefore, a technical scheme capable of accurately calculating the heat of the resource file and processing the cache file in the CDN node is needed.

Disclosure of Invention

In view of this, the embodiment of the invention provides a method and a device for calculating heat, so as to solve the problem of inaccurate heat calculation.

In order to achieve the above object, the embodiment of the present invention provides the following technical solutions:

a method of calculating heat comprising:

monitoring the number of times that a target file is requested to be accessed by a terminal and is not cached by an edge server in a preset heat calculation period;

The edge server is a cache server between the terminal and the resource server;

and if the times reach a threshold value, calculating the heat value of the target file at the current moment based on the heat calculation period.

Optionally, the calculating the heat value of the target file at the current moment based on the heat calculation period includes:

determining a first heat calculation period at the current moment and a second heat calculation period which is before and adjacent to the first heat calculation period;

obtaining a first heat value of the target file in the first heat calculation period;

obtaining a second heat value of the target file in the second heat calculation period;

and determining the heat value of the target file at the current moment based on the first heat value and the second heat value.

Optionally, the obtaining the first heat value of the target file in the first heat calculation period includes:

and based on the times, obtaining a first heat value of the target file in the first heat calculation period.

Optionally, the obtaining, based on the number of times, a first heat value of the target file in the first heat calculation period includes:

Obtaining a first heat value of the target file in the first heat calculation period by using f (HV) _1= (TM/tc_1) ×n_1;

wherein TM is the heat calculation period, and tc_1 is a time period from a time when the target file is accessed for the first time to a current time in the first heat calculation period; n_1 is the number of times the target file is requested to be accessed by the terminal and is not cached by the edge server in the period of TC_1; f (HV) _1 is a first heat value of the target file within the first heat calculation period.

Optionally, the obtaining a second heat value of the target file in the second heat calculation period includes:

obtaining a second heat value of the target file in the second heat calculation period by using f (HV) _2= (TM/tc_2) ×n_2;

wherein TM is the period of the heat calculation period, tc_2 is the period duration of the second heat calculation period or the period between the time when the target file is accessed for the first time and the end time of the second heat calculation period; n_2 is the number of times the target file is requested to be accessed by the terminal and is not cached by the edge server in the period of TC_2; f (HV) _2 is a second heat value of the target file within the second heat calculation period.

Optionally, after calculating the heat value of the target file at the current moment, the method further includes:

judging whether the heat value of the target file is larger than or equal to a preset heat threshold value;

and if the heat value of the target file is greater than or equal to the heat threshold, processing the cached file in the edge server based on the target file.

Optionally, the processing the cached file in the edge server based on the target file includes:

judging whether the size of the residual storage space of the edge server is larger than or equal to the size of the storage space required by the target file;

if the size of the residual storage space of the edge server is larger than or equal to the size of the storage space required by the target file, storing the target file to the edge server;

and if the size of the residual storage space of the edge server is smaller than the size of the storage space required by the target file, obtaining the current heat value of all the cached files in the edge server, comparing the heat value of the target file with the lowest heat value in the edge server, if the heat value of the target file is larger than the lowest heat value in the edge server, deleting the cached file with the lowest heat value in the edge server, and storing the target file to the edge server.

Optionally, after storing the target file in the edge server, the method further includes:

and setting the lowest heat value in the current heat values in the edge server as a new heat threshold.

An apparatus for calculating heat, comprising:

the monitoring unit is used for monitoring the times that the target file is requested to be accessed by the terminal and is not cached by the edge server in a preset heat calculation period; the edge server is a cache server between the terminal and the resource server;

and the calculating unit is used for calculating the heat value of the target file at the current moment based on the heat calculation period if the times reach a threshold value.

Optionally, after the calculating unit, the method further includes:

the judging unit is used for judging whether the heat value of the target file is larger than or equal to a preset heat threshold value;

and the processing unit is used for processing the cached file in the edge server based on the target file if the heat value of the target file is greater than or equal to the heat threshold.

According to the method and the device for calculating the heat, when the heat value of the target file which is not stored in the edge server is calculated, the heat calculation period is set, the number of times the target file in the heat calculation period is requested to be accessed is monitored, and when the number of times reaches the threshold value, the heat value of the target file at the current moment is calculated based on the heat calculation period, so that heat calculation is achieved. Therefore, different from the scheme of calculating the heat value through the quotient of the total access quantity and the accumulated time in the prior art, the heat calculation is set with a heat calculation period in the method, so that the calculated heat value is closer to the actual heat at the current moment, and the accuracy of heat calculation is improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flowchart of a method for calculating heat according to an embodiment of the present invention;

fig. 2 and fig. 3 are application example diagrams of a method for calculating heat according to another embodiment of the invention;

FIG. 4 is a flowchart of a method for calculating heat according to another embodiment of the present invention;

FIGS. 5-8 are diagrams illustrating another application example of a method for calculating heat according to another embodiment of the present invention;

FIGS. 9 and 10 are flowcharts illustrating a method for calculating heat according to another embodiment of the present invention;

fig. 11 and fig. 12 are schematic diagrams of an apparatus for calculating heat according to another embodiment of the invention;

fig. 13-18 are respectively diagrams illustrating another exemplary method for calculating heat according to another embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but 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.

The method for calculating heat provided by the embodiment of the invention, as shown in fig. 1, comprises the following steps:

s101, monitoring the number of times that the target file is requested to be accessed by the terminal and is not cached by the edge server in a preset heat calculation period.

The edge server is a cache server between the terminal and the resource server.

The heat calculation period may be a preset interval, as shown in fig. 2, for example, the interval of the heat calculation period may be 2 hours, 1.5 hours, or the like, as one heat calculation period in 13:00 to 14:00.

It should be noted that, when each heat calculation period is finished, the number of times that the target file obtained in the period is requested to be accessed by the terminal and is not cached by the edge server is cleared, and counting is restarted when the next heat calculation period is started.

Specifically, the starting time and the ending time of the heat calculation period may be fixed whole-point time, or may be set according to the user requirement, as shown in fig. 3, in one heat calculation period, for example, 13:00 to 14: in the period of 00, the target file can be a new file which is just online, and can be online at the speed of 13:00, and the time that the target file is requested to be accessed by the terminal and is not cached by the edge server can be directly monitored; or the target file can be on line in a 13:20 way, and the target file is monitored from the 13:20 way, the number of times the terminal requests to access and is not cached by the edge server, and the last ending time is reached; the method can also be 12:50 online, because at the end of each heat calculation period, the number of times that the target file obtained in the period is requested to be accessed by the terminal and is not cached by the edge server is cleared, and then the number of times that the target file is requested to be accessed by the terminal and is not cached by the edge server from 13:00 is only needed to be obtained.

S102, if the number of times that the target file is requested to be accessed by the terminal and is not cached by the edge server in the heat calculation period reaches a threshold value, calculating the heat value of the target file at the current moment based on the heat calculation period.

Wherein the threshold is a value used to determine whether the target file reaches a heat level at which heat calculation can be performed, for example: when the threshold value is 600 and the number of times that the target file is requested to be accessed by the terminal and is not cached by the edge server in the heat calculation period reaches 600, the heat value of the target file at the current moment can be calculated based on the heat calculation period.

It should be noted that the threshold and the heat calculation period may be dynamically adjusted according to the heat value of the cached file in the edge server.

Specifically, the threshold may be adjusted according to the popularity value of the cached file in the edge server, for example, the popularity value of the cached file in the edge server is already high, so that the threshold may be raised at this time, and the files with higher relative popularity are screened from the beginning by the number of times the target file is requested to be accessed by the terminal and not cached by the edge server in the preset popularity calculation period, and then the popularity value is calculated.

In addition, the heat calculation period can be determined by comparing the calculated current heat value with the heat value in the historical data. Specifically, if the change amplitude of the heat value is large, for example, the heat value of the current heat calculation period is 5000, and the heat value of the historical heat calculation period is 100, which indicates that the heat value of the target file is very unstable, the interval of the heat calculation periods should be properly reduced, and the timeliness of the heat value of the cached file in the edge server is ensured through more times of heat value calculation; if the change amplitude of the heat value is smaller, for example, the heat value of the current heat calculation period is 5000, and the heat value of the historical heat calculation period is 4900, which indicates that the heat value of the target file is relatively stable, the heat calculation period can be properly adjusted, the number of times of heat value calculation is reduced, and the load of the server is reduced.

As can be seen from the foregoing solutions, in the method for calculating heat according to the embodiments of the present application, when a heat value of a target file that is not stored in an edge server is calculated, a heat calculation period is set, and then, the number of times that the target file is requested to be accessed in the heat calculation period is monitored, and when the number of times reaches a threshold value, the heat value of the target file at the current time is calculated based on the heat calculation period, so that heat calculation is implemented. Therefore, different from the scheme of calculating the heat value through the quotient of the total access quantity and the accumulated time in the prior art, the heat calculation is set with a heat calculation period in the method, so that the calculated heat value is closer to the actual heat at the current moment, and the accuracy of heat calculation is improved.

Optionally, in another embodiment of the present invention, an implementation manner of the step S102 is specifically shown in fig. 4:

s401, determining a first heat calculation period in which the current moment is located and a second heat calculation period which is located before and adjacent to the first heat calculation period.

As shown in fig. 5, the first heat calculation period is the whole heat calculation period in which the trigger time TC of the secondary heat calculation is located, i.e., within a period of T2-T3; the second heat calculation cycle is a heat calculation cycle preceding and adjacent to the first heat calculation cycle, i.e., within the period of time T1-T2.

S402, obtaining a first heat value of the target file in a first heat calculation period.

Optionally, in another embodiment of the present invention, obtaining the first popularity value of the target file in the first popularity calculation period is based on the number of times the target file is requested to be accessed by the terminal and not cached by the edge server in the popularity calculation period.

When the number of times that the target file is requested to be accessed by the terminal and is not cached by the edge server in the heat calculation period reaches a preset threshold, the current heat value of the target file can be calculated according to the heat calculation period, the heat calculation trigger time and the calculation formula.

Optionally, in another embodiment of the present invention, an implementation manner of the step S402 is specifically as follows:

obtaining a first heat value of the target file in a first heat calculation period by using f (HV) _1= (TM/TC_1) ×N_1;

wherein, TM is a heat calculation period, and TC_1 is the time length from the first time when the target file is accessed to the current time in the first heat calculation period; n_1 is the number of times the target file is requested to be accessed by the terminal and not cached by the edge server in the TC_1 period; f (HV) _1 is the first heat value of the target file in the first heat calculation cycle.

Specifically, when the time TS at which the target file is accessed for the first time and the heat calculation trigger time (current time) TC are the times shown in fig. 6, the first heat value may be calculated by: according to the number of times N_1 that the target file is requested to be accessed by the terminal and is not cached by the edge server, the duration from the time TS when the target file is accessed for the first time to the hot spot calculation triggering time TC in the first heat calculation period, namely TC_1=TC-TS and the heat calculation period TM, and the first heat value of the target file at the current time is calculated by using a formula f (HV) _1= (TM/TC_1) and N_1.

When the time TS at which the target file is accessed for the first time and the heat calculation trigger time TC are the times shown in fig. 7, the first heat value may be calculated by: at the end of each heat calculation period, the number of times that the target file obtained in the period is requested to be accessed by the terminal and is not cached by the edge server is cleared, so that tc_1=tc-T3, and the formula f (HV) _1= (TM/tc_1) ×n_1 is utilized to calculate the first heat value of the target file at the current moment.

As can be seen, if the first time when the target file is accessed and the trigger time of heat calculation are within the first heat calculation period, tc_1 is the time period from the first time when the target file is accessed to the trigger time of heat calculation; if the first accessed time and the trigger time of the heat calculation are in different heat calculation periods, the number of times that the target file obtained in the period is requested to be accessed by the terminal and is not cached by the edge server is cleared when each heat calculation period is ended, and the counting is restarted when the next heat calculation period is started, wherein the first accessed time of the target file is always the time period TC_1 from the starting time of the first heat calculation period in which the trigger time of the heat calculation is located to the trigger time of the heat calculation. And then the first heat value f (HV) _1 of the target file is calculated by monitoring the number of times N_1 that the target file is requested to be accessed by the terminal and is not cached by the edge server in a preset heat calculation period and a preset heat calculation period TM.

S403, obtaining a second heat value of the target file in a second heat calculation period.

Optionally, in another embodiment of the present invention, an implementation of the step includes:

obtaining a second heat value of the target file in a second heat calculation period by using f (HV) _2= (TM/TC_2) ×N_2;

wherein TM is a heat calculation period, tc_2 is a period duration of a second heat calculation period or a period duration between a time when the target file is accessed for the first time and an end time of the second heat calculation period; n_2 is the number of times the target file is requested to be accessed by the terminal and not cached by the edge server in the period of TC_2; f (HV) _2 is the second heat value of the target file in the second heat calculation cycle.

Specifically, when the time TS at which the target file is first accessed and the heat calculation trigger time TC are the times shown in fig. 6, since the time TS at which the target file is first accessed does not pass through the end of any one heat calculation cycle, there is no heat value record about the target file in the edge server, so tc— 2=0.

When the time TS at which the target file is accessed for the first time and the heat calculation trigger time TC are the times shown in fig. 7, the second heat value may be calculated by: when TC is in the first heat calculation period and TS is before the second heat calculation period, the second heat value is the heat value of the target file in the whole second heat calculation period, at this time, tc_2=t3-t2, n_2 is also the number of times the target file monitored in the whole second heat calculation period is requested to be accessed by the terminal and is not cached by the edge server, and the first heat value of the target file at the current moment is calculated by using formula f (HV) _2= (TM/tc_2) ×n_2.

When the time TS at which the target file is accessed for the first time and the heat calculation trigger time TC are the times shown in fig. 8, the second heat value may be calculated by: when TC is in the first heat calculation period and TS is in the second heat calculation period, at this time, tc_2=t3—ts, and the first heat value of the target file at the current time is calculated by using formula f (HV) _2= (TM/tc_2) ×n_2.

It should be noted that, because the second heat calculation period is one heat calculation period before and adjacent to the first heat calculation period, if the first time when the target file is accessed is within the second heat calculation period, tc_2 is a time period between the first time when the target file is accessed and the end time of the second heat calculation period; if the first accessed time of the target file is at or before the starting time of the second heat calculation period, TC_2 is the period duration of the second heat calculation period, and then the second heat value f (HV) _2 of the target file is calculated by monitoring the number of times N_2 that the target file is requested to be accessed by the terminal and is not cached by the edge server in the preset second heat calculation period and the preset heat calculation period TM; if the first time the target file is accessed is in the first hot period, f (HV) _2=0.

S404, determining the heat value of the target file at the current moment based on the first heat value and the second heat value.

Comparing the obtained first heat value with the second heat value, wherein the heat value of the target file at the current moment takes a larger value, for example, the first heat value is larger than the second heat value, and then the first heat value is used as the heat value of the target file at the current moment; if the first heat value is smaller than the second heat value, taking the second heat value as the heat value of the target file at the current moment; and if the first heat value is equal to the second heat value, the heat value of the current moment of the target file is the first heat value or the second heat value.

Optionally, in another embodiment of the present invention, as shown in fig. 9, after calculating the heat value of the target file at the current time, the method further includes:

s901, judging whether the heat value of the target file is larger than or equal to a preset heat threshold, and if the heat value of the target file is larger than or equal to the heat threshold, executing step S902.

Wherein the heat threshold is a value for comparing with the heat value of the target file at the current time, and if the current heat value of the target file reaches the heat threshold, step S902 may be executed.

S902, processing the cached files in the edge server based on the target file.

It should be noted that, if the popularity value of the target file is smaller than the popularity threshold, the process returns to step S101 to continuously monitor the number of times that the target file is requested to be accessed by the terminal and is not cached by the edge server in the popularity calculation period.

Optionally, in another embodiment of the present invention, step S902 of processing the cached file in the edge server based on the target file may specifically be: storing the target file into an edge server, and providing the target file as a cached file for a terminal to access. Specifically, in one embodiment of step S902, as shown in fig. 10, the following steps may be implemented:

s1001, judging whether the size of the residual storage space of the edge server is larger than or equal to the size of the storage space required by the target file.

It should be noted that, if the size of the remaining storage space of the edge server is greater than or equal to the size of the storage space required by the target file, step S1002 is performed; if the size of the remaining storage space of the edge server is smaller than the size of the storage space required for the target file, step S1003 is performed.

S1002, storing the target file to the edge server.

S1003, obtaining the current heat value of all cached files in the edge server, and executing step S1004.

Acquiring a first heat value of a cached file in a first heat calculation period; and obtaining a second heat value of the cached file in the second heat calculation period, and determining the heat value of the cached file at the current moment based on the first heat value and the second heat value.

It should be noted that, the current heat value of the cached file is obtained by calculating the heat value. The method for calculating the heat value can refer to the method content in step S102 and the corresponding embodiment content, which are not described herein.

S1004, comparing the lowest heat value in the edge server with the heat value of the target file.

It should be noted that, if the heat value of the target file is greater than the lowest heat value in the edge server, step S1005 is executed; if the target file has a heat value less than the lowest heat value in the edge server, the cached file continues to be maintained.

S1005, deleting the cached file with the lowest heat value in the edge server, and storing the target file in the edge server.

The target file is a file that is not cached in the edge server, so when the popularity value is equal to the popularity value of the cached file in the edge server, the target file is preferentially stored in the edge server, and the cached file with the popularity value equal to the target file is deleted.

It should be noted that in the embodiment, the files with low heat value in the edge server may be further cleaned at regular time when the heat value in the edge server is lower than that of the cached files, specifically, in the embodiment, the cached files in the edge server are sorted through heat value calculation, and the files with low heat value are sequentially deleted, so that the remaining space in the edge server meets the remaining space threshold, and enough space in the edge server is ensured to be used for coping with file injection under emergency. The threshold of the remaining space may be 80% of the total storage space of the edge server, or any percentage size.

Optionally, in another embodiment of the present invention, after storing the target file in the edge server, the method further includes:

the lowest heat value of the current heat values in the edge server is set as the new heat threshold.

It should be noted that if no target file reaches the heat threshold, the target file needs to be stored in the edge server, and each time the heat calculation period is finished, all the cached files in the edge server are calculated and ordered once, and the lowest heat value in the cached files is set as a new heat value.

The embodiment of the invention also provides a device for calculating heat, as shown in fig. 11, which comprises:

the monitoring unit 1101 is configured to monitor the number of times that the target file is requested to be accessed by the terminal and is not cached by the edge server in a preset heat calculation period; the edge server is a cache server between the terminal and the resource server.

And the calculating unit 1102 is configured to calculate, based on the heat calculation period, a heat value of the target file at a current time, where the number of times that the target file is requested to be accessed by the terminal and is not cached by the edge server in the heat calculation period reaches a threshold.

In the device for calculating heat disclosed in this embodiment, the specific process of each unit may refer to the content of the method embodiment corresponding to fig. 1, which is not described herein again.

Optionally, in another embodiment of the present invention, after the computing unit 1102, as shown in fig. 12, the method further includes:

a determining unit 1201, configured to determine whether the heat value of the target file is greater than or equal to a preset heat threshold.

A processing unit 1202, configured to process the cached file in the edge server based on the target file if the popularity value of the target file is greater than or equal to the popularity threshold.

In this embodiment, the specific execution of the determining unit 1201 and the processing unit 1202 may refer to the method embodiment corresponding to fig. 9, and will not be described herein.

As can be seen from the foregoing, in the device for calculating heat provided by the embodiment of the present application, when calculating the heat value of the target file that is not stored in the edge server, the heat calculation period is set, and then the number of times that the target file is requested to be accessed in the heat calculation period is monitored, and when the number of times reaches the threshold value, the heat value of the target file at the current moment is calculated based on the heat calculation period, so as to implement heat calculation. Therefore, different from the scheme of calculating the heat value through the quotient of the total access quantity and the accumulated time in the prior art, the heat calculation is set with a heat calculation period in the method, so that the calculated heat value is closer to the actual heat at the current moment, and the accuracy of heat calculation is improved.

For the above implementation of this embodiment, the following exemplifies a specific implementation of this embodiment:

in connection with the scenario specifically implemented in this embodiment shown in fig. 13, when the edge server is requested to be accessed by a user and there is no file requested by the user cached in the edge server, the number of times N that the content file is requested to be accessed by the terminal and is not cached by the edge server is recorded, and when N reaches the set threshold M and meets the minimum heat value of the existing cache at the same time, a downloading procedure of the file is started, and the file is downloaded from the central node to meet the user requirement.

The heat value calculation theory of the embodiment can realize the calculation mode under the condition of different time by the following scheme:

1. calculated according to a heat calculation cycle, for example, from 13:00 to 14:00 as one heat calculation cycle.

2. The current popularity value of the cached content consists of a historical popularity value and a current popularity value, which are maximized.

3. The historical heat value takes the current time as a reference, and a heat calculation period section is shifted upwards as the heat value of the file, for example: the current time is 14:20, then the actual heat calculation period of the historical heat value, the second heat calculation period, starts from 13:00 to 14:00.

4. The current heat value is converted into a heat value of a file in the whole time period by taking the current time as a reference, for example: the current time is 14:20, then a statistical period of heat values is expected, i.e., the first calculation period starts from 14:00 to 15:00. In practice, only 20 minutes have elapsed, but the heat value can be converted into 1 hour as the heat value of the whole period.

The access times of the files passing through the current edge server are counted, namely: when the number of times that the terminal requests access and is not cached by the edge server reaches N times, the heat value HV is calculated, compared with the heat threshold value HM, when the current file HV value is larger than the heat threshold value HM under the condition that the storage space is sufficient, the file is cached and downloaded unconditionally and stored.

In addition, in the case that the storage space of the edge server is insufficient or the remaining storage threshold size is reached, the following scheme may be adopted in this embodiment:

1. calculating the residual storage capacity space of the edge server and marking as S1;

2. the size of the file that needs to be cached is denoted S2.

3. And comparing the S1 value with the S2 value, and when the S2 value is larger than the S1 value, taking out a file with the lowest cached heat value from the current cached file library, and deleting the file first.

4. Confirming the current storage space size S1 (updated residual space value), if S1 is greater than S2, executing the 5 th step, otherwise returning to the 3 rd step, and continuously deleting the file with the lowest heat value.

5. And caching the new file.

And comparing the heat value HV of the currently injected file with the heat value of the currently cached file, and if the minimum value of the heat values of the cached file is larger than the HV value, adjusting the heat threshold HM to be equal to the final heat value HV.

Some attention is paid to: total number of accesses requested by the user = total number of misses + total number of hits.

For example, with respect to the current cluster, i.e. the file is not found on the edge server and is counted as a miss once, the number of times of the file miss is equal to the total number of times of the file access; the other type is a file cached in the storage disk in the edge server, and the access times are equal to the total access times of the file.

Therefore, the calculation of the heat value of the resource aims at all resource requests passing through the edge server, no matter the miss and hit, the total access times of the user requests are taken as the basis, and the total access times are used as a whole calculation for calculation and sequencing, so that the fairness of resource access is met, and the data of the heat calculation is more objective and accurate.

It should be noted that, the dynamic adjustment of the heat threshold HM is to reduce the number of fast-forwarding and fast-out times of the cached file in the edge server, and when the heat calculation time arrives, the minimum heat value of the file cached in the edge disk is reassigned to the heat threshold M, but cannot be smaller than the heat threshold set by default. I.e. there is a lower limit but no upper limit.

Specifically, the detailed calculation scheme of the heat value in this embodiment is as follows:

1. the calculation formula of the file heat value comprises the following steps:

and (3) calculating the access times N of the file in a certain heat calculation period TM value for the file of the source, and triggering a heat judgment flow when the access times threshold HM is reached. The time length triggered from the first access to the number of accesses is also timed TC.

The source returning refers to returning to the source station server, and in this embodiment, the source station server is a storage server of the central node.

When the time TM value is not reached, the number of accesses N needs to be estimated as the access heat value of the file within the entire TM value.

The TM value reaches a threshold value when the TM value is tentatively updated for 1 hour, the whole point per hour, and the threshold value of the default initial access number is 3 times/minute, that is, the number of times the TM value is accessed for 1 hour=60×3=180 times, and the HM value is calculated from f (HV).

It should be noted that, the TM value and the access number threshold may be dynamically adjustable, and the calculation period of heat is 1 hour for convenience of description of the present invention.

The formula f (HV) is as follows:

the condition is HV > =hm, the heat value takes integer form, and the creation time is compared if the heat value is the same.

Wherein, the heat value rounding means that only an integer is left, and the positive rounding means that decimal points are removed. The negative number is rounded up to take the largest integer not greater than this negative number.

It should be noted that when the files are cached in the edge server, there is a time for creating and storing the files in the edge server, and when the storage space is insufficient and the heat of the two files is the lowest and the same, then the files with earlier creation times need to be deleted preferentially.

Access heat value f (HV) =tm (N/TC) = (TM/TC) ×n;

if accessed 1000 times within tc=10 minutes, then tm=60 minutes accesses the heat value:

f (HV) =60 (1000/10) =6000 (heat value).

That is, in a certain time range, the higher the access times, the larger the heat value of the file, and the two are in direct proportion.

2. Calculating the heat value of the uncached file:

when the triggering time of each heat calculation starts and is one integral point time, actively triggering to update all the file heat values accessed by the edge server once, and directly calculating according to a formula if the interval time of the file heat values meets the TM value. When the interval time is less than the TM value, the current elapsed time period is converted with the TM value, and the less TM value is estimated into the heat value in the TM time period.

3. Calculating and sequencing the heat value of the cached file:

in this embodiment, the heat value of each file of each time TM value is counted and sorted mainly for the files cached in the edge server.

4. The processing scheme when the storage space of the edge server is insufficient is as follows:

and calculating the heat value of all files in the edge server, and selecting the file with the lowest rank but lower heat value than the file to be injected according to the size and the heat value of the file to be injected, and deleting the file until the deleted space meets the size of the file to be cached. And finally, updating the lowest heat value in the edge server to be a new heat threshold value.

5. Theoretical calculation model:

as shown in fig. 14, according to the number of missed files, when the number of missed files is greater than the threshold value HM, the heat determining process is triggered, and when the storage space is satisfied, the downloading process of the target file is directly triggered, otherwise, the edge server needs to be cleaned to free space, namely, the access heat value of the missed file is compared with the smallest heat value of the cached file, when the heat value of the missed file is greater than the smallest heat value of the cached file, the smallest heat value of the cached file is deleted, and the downloading process of the missed file is stopped until the space is satisfied, and when the storage space is satisfied, the downloading process of the missed file is entered.

It should be noted that, when ranking values with similar or same heat degree appear, the ranking values are first cleared by the creator based on the creation time.

Specifically, the practical application algorithm in this embodiment is as follows:

first, four values are stored for each file accessed at the edge server in this embodiment:

the actual access times during integral point calculation are marked as LAT;

the actual access value per minute during whole point calculation is marked as LCR;

the actual access number in non-integral point calculation is recorded as CAT;

the estimated access value per minute at the time of non-integral point calculation is noted as CCR.

Wherein the different variables are described as follows:

1. TS represents a start time;

2. TC represents a heat calculation trigger time;

3. TTX (X values are 1, 2, 3, 4) indicating the whole time;

in the first heat calculation case, as shown in fig. 15, when TS and TC are both within one heat calculation period, when the heat value is calculated, then the above values are respectively:

LAT value is 0;

LCR value is 0;

the CAT value is the actual access time value from TS time to TC time;

CCR values: firstly, CAT value is estimated as an estimated value from TT1 time to TT2 time, and the heat value HV, namely CCR value is obtained by referring to the calculation of formula f (HV). Both LCR and CCR values are prioritized, delaying the downward trend.

f(HV)_1＝0；

f(HV)_2＝(TM/(TC-TS))*N；

f (HV) =max (f (HV) _1, f (HV) _2) takes the high score.

In the second heat calculation case, as shown in fig. 16, when TC is in the first heat calculation period and TS is in the second heat calculation period, and when the heat values are calculated, then the above values are respectively:

the LAT value is the actual access times value from TS time to TT2 time;

the LCR value is the actual access value from TS moment to TT2 moment, and the heat value is obtained by calculation according to a formula f (HV);

the CAT value is the actual access time value from the TT2 time to the TC time;

CCR values: first, the CAT value is estimated as an estimated value from the time TT2 to the time TT3, and the HV value, i.e., the CCR value, is calculated with reference to the formula f (HV). Both LCR and CCR values are prioritized, delaying the downward trend.

f(HV)_1＝(TM/(TT2-TS))*N；

f(HV)_2＝(TM/(TC-TT2))*N；

f (HV) =max (f (HV) _1, f (HV) _2) takes the high score.

In the third heat calculation case, as shown in fig. 17, when TC is in the first heat calculation period and TS is before the second heat calculation period, the calculation time changes, and all values change. The actual heat value needs to be calculated in the range from TT2 to TC, so when calculating the heat value, the above values are respectively:

the LAT value is the actual access times value from TT2 to TT 3;

the LCR value is the actual access value from the TT2 time to the TT3 time, and the heat value is obtained through calculation according to a formula f (HV);

The CAT value is the actual access time value from the TT3 time to the TC time;

CCR values: firstly, CAT value is estimated as an estimated value from TT3 time to TT4 time, and HV value, namely CCR value, is obtained by calculation with reference to formula f (HV). Both LCR and CCR values are prioritized, delaying the downward trend.

f(HV)_1＝T2；

f(HV)_2＝(TM/(TC-TT3))*N；

f (HV) =max (f (HV) _1, f (HV) _2) takes the high score.

According to the analysis of the above situation, when the calculation time TC is within the calculation time period TM, that is, the heat value HV thereof is estimated according to time, the final value thereof takes the estimated value thereof and compares the heat value in the previous calculation time period with the higher score. When the calculation period full point is reached, that is, (TT 1, TT2, TT3.. TTn) in the figure, the heat value thereof is calculated from the actual number of accesses. I.e. the heat value is always taken closer to being calculated with the real access amount and time.

The heat value after the whole point is calculated as the actual access times and the elapsed time, and the access times are 0.

Referring to fig. 15, the active disk space clearing logic in this embodiment is as follows:

the disk space active clearing logic in this embodiment has two deletion methods:

1. real-time storage purge

And judging whether enough space exists in the residual storage space to store the file according to the file heat value to be stored, and if the file is insufficient, cleaning out a proper space value according to the heat value, wherein the file size value and the creation time are generally used as the reference.

2. Timed store purge

And cleaning out enough disk space of the edge server according to the residual space threshold at a certain moment in time for coping with the emergency.

For example, the current edge server is fixed at the 24 am point 00 time sharing to sort the heat values of all the cached files, when the edge server is sufficient in space, no operation is performed, when the storage space is deficient, in order to meet the injection requirement of the subsequent files, intelligent deletion is performed according to the empty preset value of the capacity of the edge server, and the deletion is performed from small to large according to the heat value until the free available capacity reaches the expected space value.

There are two thresholds: for example, the total space for injecting the cache file data is total space, and in operation, the upper limit storage space of the cache can be set to be 85%, that is, the total space is 85%, the total space is not needed to be full, and the possible disk fragmentation problem is considered, and a certain redundancy can be reserved for key information storage with important points, such as the cached file list information (in order to prevent the database server from normally providing service when not working). In order to make the capacity of the storage space reserve enough space, the standard storage upper limit space can be set to be 80%, and the standard storage upper limit space is only used for actively cleaning the cached files, namely, the total space is 80%.

In summary, in the two methods of the disk space active clearing logic in this embodiment, real-time clearing is focused on clearing only the appropriate space value of the media file and completing as soon as possible, and timing clearing is focused on clearing enough space for media file injection in emergency, so as to reduce the return source of media request access.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of calculating heat comprising:

the edge server is a cache server between the terminal and the resource server;

And if the times reach a threshold value, calculating the heat value of the target file at the current moment based on the heat calculation period, wherein the threshold value and the heat calculation period are dynamically adjusted according to the heat value of the cached file in the edge server.

2. The method of claim 1, wherein calculating a heat value of the target file at a current time based on the heat calculation period comprises:

3. The method of claim 2, wherein the obtaining a first heat value of the target file during the first heat calculation period comprises:

4. A method according to claim 3, wherein said obtaining a first heat value of said target file within said first heat calculation period based on said number of times comprises:

5. The method of claim 2, wherein the obtaining a second heat value of the target file within the second heat calculation period comprises:

6. The method of claim 1, further comprising, after calculating the heat value of the target file at the current time,:

7. The method of claim 6, wherein the processing the cached file in the edge server based on the target file comprises:

8. The method of claim 7, further comprising, after storing the target file to the edge server:

9. An apparatus for calculating heat, comprising:

and the calculation unit is used for calculating the heat value of the target file at the current moment based on the heat calculation period if the times reach a threshold value, wherein the threshold value and the heat calculation period are dynamically adjusted according to the heat value of the cached file in the edge server.

10. The apparatus of claim 9, wherein the computing unit is further configured to: