JP2010224647A - Timer value management system and server device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a system for dynamically setting a completion timer value of a device of a higher layer in an information processing system having a multistage structure of three layers or more. <P>SOLUTION: The system includes an information collection means provided on each server included in a multistage server system having three or more layers, the information collection means collecting completion time information showing, for a processing request transmitted to a server lower in order than the own server, the time to completion of processing by servers located in layers up to the server of the transmitting destination; a completion timer value table provided on each server, the table holding a completion timer value based on the collected completion time information in conformation to the transmitted processing request, and being used for time-out control for the processing request; and a report means provided on each server, the report means determining a completion time including the processing time of the own device based on the collected completion time information and the time required for processing in the own device, and transmits a completion time report containing completion time information showing this completion time to a server located in a layer higher in order than the own device. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a timer value management system and a server device having a timer value management function in a multistage server system.

  For example, a Web system generally has a configuration including three layers: a Web server positioned at the highest level, an application server positioned lower than the Web server, and a database server at the lowest level. The network operation system also generally has a three-layer configuration in which an application server and a network device are located below a top layer front server.

  In an information processing system having such a hierarchical structure, processing requested by servers located in the middle layer and the lower layer is executed in response to a request from the upper layer, and the processing result is sequentially obtained upon completion of the processing. A response containing is returned to the upper layer server. In addition to such a sequential response type information processing system, there is also a polling type information processing system in which an upper layer server performs polling and collects processing results from the middle layer and hierarchy servers.

  In the sequential response type information processing system described above, based on whether or not the time from when the request is sent until the response from the lower-layer server is returned is usually within a predetermined time set in the timer, It is determined whether or not the lower layer server is operating normally. On the other hand, in a polling type information processing system, a timer in which an estimated required time for a requested process is set is used to manage the timing at which an upper layer server performs polling (see Patent Document 1).

  Conventionally, in any type of information processing system, the timer value set in the timer is generally determined by the person in charge of managing the information processing system, for example, when the information processing system is constructed. . At this time, the person in charge of system management refers to the specifications, etc., and the time required for each process to be completed based on the processing capacity of the middle and lower tier servers and the expected load. And the completion timer value to be set in the timer is determined.

JP-A-4-239341

  Incidentally, in recent years, an information processing system as described above is configured by combining server products of various specifications made by various manufacturers. Further, in the system expansion, a server with a manufacturer and specification different from the server already incorporated in the information processing system may be added.

  For this reason, it is a very heavy burden on the person in charge of system management to estimate an appropriate time required for processing corresponding to each request at the time of system construction and to determine an appropriate completion timer value. Further, every time the system configuration is changed, it is necessary to change the timer value set in the changed hierarchy and the server positioned higher than that, and the work is complicated.

  It is an object of the present invention to provide a timer value management system and a server device having a mechanism for dynamically setting a completion timer value of an upper layer device in an information processing system having a multistage configuration of three or more layers. .

  The above-described object can be achieved by a timer value management system disclosed below.

  A timer value management system according to a first aspect is provided in each server included in a multistage server system having three or more layers, and is a destination server for processing requests sent to a server lower than its own device. Information collection means for collecting completion time information indicating the time until processing by the servers located in the previous hierarchy is completed, and each server is provided with the collected completion time information corresponding to the transmitted processing request. Based on the completion timer value table that holds the completion timer value based on the completion time value for the processing request and the collected completion time information and the time required for the processing in the own device. The completion time including the time is obtained, and the completion time report including the completion time information indicating the completion time is transmitted to the server located at a higher level than the own device. And a reporting means.

  In the timer value management system configured in this way, for example, when a processing request from the upper server is relayed to the lower server via the intermediate server, the reporting means provided in the lower server by the collecting means of the intermediate server Travel time reports from are collected. At this time, based on the collected completion time information, the report means obtains the completion time until the processing up to the intermediate server is completed, and reports it to the upper server. Upon receipt of the completion time report from the intermediate server, the completion timer value table is updated based on the completion time information collected by the collection means of the upper server, and the updated completion timer value is subjected to timeout control.

  According to the timer value management system having the configuration described above, the completion time until the completion of the processing up to the server of the hierarchy located lower than the own device is collected in the server of each hierarchy, and the collected completion time is Based on this, the completion timer value table is updated. That is, in an information processing system having a multi-level configuration of three or more layers, it is possible to realize a mechanism for dynamically setting a completion timer value for processing up to a layer lower than the own device in the server device of each layer. .

  As a result, an appropriate completion timer value can be set in the completion timer value table in accordance with the change or addition of the model of the server device included in each layer, so that a complicated information processing system including a large number of layers can be managed. Can be freed from complicated timer value management.

It is a figure which shows one Embodiment of a timer value management system. It is a figure which shows the detailed structure of a completion time calculation part. It is a figure which shows the example of each table. It is a flowchart showing the operation | movement at the time of processing request reception. It is a flowchart showing the operation | movement at the time of completion time report reception. It is a sequence diagram explaining the operation | movement which manages a completion timer value. It is a figure which shows another embodiment of a timer value management system. It is a figure which shows another embodiment of a completion timer value table. It is a sequence diagram explaining the operation | movement which manages a completion timer value.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
(Embodiment 1)
FIG. 1 shows an embodiment of a timer value management system.

  The processing request sent from the processing control unit 211 of the upper server 210 shown in FIG. 1 is relayed to the processing control unit 231 of the lower server 230 via the processing control unit 221 of the intermediate server 220. In addition, a response including a result obtained by executing the process specified by the process request by the process control unit 231 of the lower server 230 is sent to the process control of the upper server 210 via the process control unit 221 of the intermediate server 220. Returned to the part 211.

  In such a three-tier server system, the timeout control units 212 and 222 provided in the higher-level server 210 and the intermediate server 220 perform processing requests based on the completion timer values registered in the completion timer value tables 213 and 223. Timeout control is performed every time.

  In the completion timer value tables 213 and 223, as will be described later, the completion timer value indicating the completion time required for completing the process up to the lower layer than the own device is stored in the type of the processing request and the request destination. Registered corresponding to the combination with the server.

FIG. 3B shows an example of a completion timer value table. In this example, respective completion timer values are stored corresponding to combinations of request types A, B, and C and servers indicated by server IDs (S ₁₀ ), (S ₂₀ ), and (S ₃₀ ). Yes. Here, the server ID of the request destination is a server ID indicating the intermediate server 220 in the completion timer value table 213 provided in the upper server 210. On the other hand, in the completion timer value table 223 provided in the intermediate server 220, the server ID indicates the lower server 230. In the initial state of the timer management system, the completion timer value tables 213 and 223 can hold initial values corresponding to the respective combinations. Further, the processing is started with the completion timer value tables 213 and 223 being emptied, and default values can be registered as will be described later.

  In addition, the completion time calculation unit 224 provided in the intermediate server 220 determines that the processing up to the own device is completed based on the completion timer value registered in the completion timer value table 223 and the processing time in the own device. Calculate the required completion time. The completion time obtained by the completion time calculation unit 224 is sent to the upper server 210 via the completion time report unit 226, and is reflected in the completion timer value table 213 by the report collection unit 214 provided in the upper server 210. Is done.

  On the other hand, the completion time calculation unit 232 provided in the lower server 230 located in the lowest hierarchy calculates the completion time until the processing in the own device is completed. The completion time obtained by the completion time calculation unit 232 is sent to the intermediate server 220 via the completion time report unit 233, and is reflected in the completion timer value table 223 by the report collection unit 225 provided in the intermediate server. The At this time, the completion time calculation unit 224 recalculates the completion time up to the own device based on the completion time collected by the report collection unit 225, and the completion time obtained by recalculation is calculated as the completion time. The report unit 226 provides a report process to the upper server 210.

  Further, in the server information tables 215, 227, and 234 shown in FIG. 1, the processing control units 211, 221, and 231 use the processing control units 211, 221, and 231 for each processing request to indicate information indicating the request type and the higher-level server that is the transmission source, And the request method to these lower servers are stored.

FIG. 3C shows an example of the server information table. In this example, a request type A and an upper server ID (M ₁ ) indicating the upper server 210 are held corresponding to the request ID (001). Further, a lower server ID (S ₂₀ ) indicating another lower server is held as information indicating a relay destination lower server together with a lower server ID (S ₁₀ ) indicating the lower server 230. The request method parallel indicates a processing procedure for sending processing requests in parallel to a plurality of lower servers and collecting responses from these lower servers. On the other hand, the request method serial indicates a processing procedure in which processing requests are sequentially transmitted to at least one lower server and responses are collected.

  Next, the completion time calculation process by the completion time calculation unit 224 will be described in detail.

  FIG. 2 shows a detailed configuration of the completion time calculation unit.

  2 that are the same as those shown in FIG. 1 are given the same reference numerals, and descriptions thereof are omitted.

  The processing time table 241 illustrated in FIG. 2 holds a standard processing time for executing the specified processing by the own device for each processing request type.

  FIG. 3A shows an example of the processing time table 241. FIG. 3C shows an example of the server information table 227.

  In the example of the processing time table shown in FIG. 3A, processing times of 300 ms, 600 ms, and 100 ms are held corresponding to the request types A, B, and C.

  Also, the load information collection unit 243 illustrated in FIG. 2 collects information indicating the load state of the intermediate server device 220 and supplies the collected load information to the processing of the processing time calculation unit 244. The processing time calculation unit 244 calculates the processing time reflecting the load state of the server device by multiplying the standard processing time held in the processing time table 241 by the coefficient α determined based on the load information. Then, it is used for the processing of the completion time calculation unit 245.

  The completion time calculation unit 245 is based on the information held in the completion timer value table 223 and the server information table 227 and the processing time obtained by the processing time calculation unit 244 in response to an instruction from the calculation control unit 246. Thus, the completion time to the intermediate server 220 is calculated.

  Further, the calculation control unit 246 causes the above-described completion time calculation unit 245 to calculate the completion time when a new processing request is received by the processing control unit 221. The calculation control unit 246 also causes the completion time calculation unit 245 described above to calculate the completion time even when the report collection unit 225 receives the completion time report from the lower server 230.

  The completion time calculation unit 232 of the lower server 230 can be configured similarly.

  FIG. 4 is a flowchart showing the operation when a processing request is received.

  When the process control unit 221 receives a process request (step 301), the process control unit 221 includes information including a request type, an upper server ID, a relay destination lower server ID, and a request method as information for specifying the process request. Is registered in the server information table 227 (step 302). At this time, the calculation control unit 246 refers to the server information table 227 and determines whether or not the processing request is relayed to a lower hierarchy than the own device based on the information indicating the relay destination of the processing request ( Step 303).

  Then, when there is a relay destination (affirmative determination in step 303), the calculation control unit 246 and the completion time calculation unit 245 execute the completion time calculation process surrounded by a broken line in FIG.

  First, the calculation control unit 246 refers to the completion timer value table 223, and for each lower server registered in the server information table 227, whether or not a completion timer value is registered corresponding to the request type of the accepted processing request. Is determined (step 304).

  If the determination in step 304 is affirmative, the calculation control unit 246 reads the completion timer value in the completion timer value table 223 corresponding to the lower server (step 305) and passes it to the completion time calculation unit 245. The calculation control unit 246 repeats Step 304 to Step 306 until the completion timer values are read for all lower server IDs held as relay destination servers in the server information table 227. In this way, when the completion timer values corresponding to all the lower servers are read (affirmative determination in step 306), the calculation control unit 246 exits the loop and proceeds to the next step.

  In the process of reading the completion timer value, if a subordinate server whose completion timer value is not registered is detected (No in Step 304), the completion time calculation unit 245 stores a predetermined value in the completion timer value table 223. A default value is registered (step 307). In this case, this default value is used as a completion timer value corresponding to an unregistered subordinate server in the subsequent completion time calculation processing. As the default value described above, a value indicating a sufficient time as a time required for processing by the lower server can be set.

  After the completion timer value reading for the lower server is completed as described above, the processing time calculation unit 244 calculates the processing time tp of its own device in response to an instruction from the calculation control unit 246 (step 308). . Then, based on the processing time tp and the completion time of each subordinate server, the completion time calculator 245 calculates the completion time tc of the processing up to the own device (step 309).

  At this time, the completion time calculation unit 245 first completes the processing by the lower server based on the request method shown in the server information table 227 and the completion timer value corresponding to each lower server read in step 305. The required time td is obtained. In the processing request to which the request method parallel is applied, the completion time calculation unit 245 sets the maximum one of the completion timer values corresponding to the lower servers to which the processing request is relayed in parallel as the required time td. On the other hand, in the processing request to which the request method serial is applied, the completion time calculation unit 245 obtains the sum of completion timer values corresponding to the relay destination lower servers as the required time td. Then, the completion time calculation unit 245 calculates the completion time tc by adding the processing time tp received from the processing time calculation unit 246 in step 308 to the required time td thus obtained.

  The completion time tc determined in this way is passed to the completion time reporting unit 226. In response to this, the completion time report unit 226 creates a completion time report including, for example, the request ID, the request type ID, the server ID of the device itself, and the completion time tc, and sends it to the upper server 210 (step 310). ).

  On the other hand, when the own apparatus is the final destination of the processing request as in the lower server 230 shown in FIG. 1, a negative determination is made in step 303, and steps 304 to 307 are skipped. In step 308, the processing time tp of the own apparatus calculated by the processing time calculation unit 246 is provided to the processing of the completion time reporting unit 226 as the completion time tc as it is.

  In this way, in response to the reception of the processing request, the time required for the processing up to the device itself is reported to the upper server, and the processing at the time of the processing reception ends.

  Next, the operation when the intermediate server 220 and the upper server 210 accept the completion time report will be described.

  FIG. 5 is a flowchart showing the operation when the completion time report is received.

  When the calculation control unit 246 receives the completion time report via the report collection unit 225 (step 311), the calculation control unit 246 refers to the server information table 227 to confirm that the processing request indicated by the request ID is being processed. Confirm (affirmative determination in step 312).

  Next, the calculation control unit 246 reads the corresponding completion timer value from the completion timer value table 223 based on the request type and server ID included in the completion time report, and whether the completion time included in the completion time report has changed. It is determined whether or not (step 313).

  If the completion time has changed (Yes in Step 313), the report collection unit 225 registers the completion time included in the completion time report as a new completion timer value in the completion timer value table 223 (Step 314). .

  Next, the calculation control unit 246 refers to the server information table 227 and confirms that there is an upper server to which the completion time report is to be transmitted (step 315). When the determination in step 315 is affirmative, the calculation control unit 246 causes the processing time calculation unit 244 and the completion time calculation unit 245 to execute the completion time calculation process surrounded by the broken line in FIG. 4 (step 316).

  In this way, when the intermediate server 220 receives the completion time report from the lower server 230, the completion time tc required to complete the process up to its own device is calculated again, and again by the completion time reporting unit 226. It is sent to the upper server 210 (step 317).

  On the other hand, in the upper server 210 that is the processing request transmission source, the determination in step 315 is negative, and the processing ends after the completion timer value table 213 is updated based on the completion time included in the completion time report.

  In this way, the timer value management system shown in FIG. 1 collects the completion time required for processing from the server located in the lower hierarchy to the hierarchy by each server apparatus, and reduces the processing time required for the processing of the own apparatus. In addition, a mechanism for reporting to a server located in a higher hierarchy is realized.

  By registering the completion time collected from the lower tier in the completion timer value table in each tier server, the completion timer value can be dynamically set and used for timeout control for processing up to the lower tier. .

  According to the timer value management system configured in this way, a completion timer provided in a server in a higher hierarchy than the hierarchy in which the configuration has been changed, regardless of which hierarchy in the multistage server system has been changed. Configuration changes can be reflected in the value table.

  For example, FIG. 6 shows a sequence for explaining the operation of managing the completion timer value in a multi-tier server system having four layers.

In the example shown in FIG. 6, the processing request from the server _{S 4} located at the top REQ1 is, are sequentially relayed to the server _{S 3, S 2, S 1} , the response RES1 including the processing result by the server S1, the server It returned to the server _{S 4} through _{S 2,} S _3.

In the process of exchanging the processing request and the response, the server S ₃ first executes the completion time calculation process P ₃₁ when receiving the processing request REQ1 from the server S ₄ . In the server S ₃ , the processing time t ₃ reflecting the processing load of the server S ₃ is added to the completion timer value in the server S ₃ corresponding to the sum of the processing time default values d ₂ and d ₁ of the servers S ₂ and S _1. to completion time _{(t 3 + d 2 + d} 1) is determined and reported to the server _{S 4} (see completion time reported (3-1)).

Similarly, in the servers S ₂ and S ₁ , when the processing requests REQ 2 and REQ 3 are received, completion time calculation processes P ₂₁ and P ₁₁ are executed, and completion times (t ₂ + d ₁ ) and (t ₁ ) are calculated, respectively. These completion times are reported to the servers S3 and S2 (see completion time reports (2-1) and (1-1)).

Depending on the above-mentioned completion time reported (1-1), the server S _2, completion time calculating process P ₂₂ is executed, completion time without the default values for the hierarchy to the server _{S 2 (t 2 + t 1} ) is is calculated and subjected to report processing to the server S ₃ (see completion time reported (2-2)). Upon receiving this completion time report, a completion time (t ₃ + t ₂ + t ₁ ) that does not include a default value is calculated for the layers up to the server S ₃ by the completion time calculation process P ₃₃ executed in the server S ₃ . It is subjected to report processing to the server S ₄ (see completion time reported (3-3)).

In response to this, for example, the completion timer value of the server S ₄ the uppermost first configured default value of the server corresponding in each layer d _3, d _2, from the sum of d _1, each of the server S _3, It is replaced with a completion time (t ₃ + t ₂ + t ₁ ) reflecting the processing time in S ₂ and S ₁ .

Note that the default value of the processing time by the servers S ₃ , S ₂ , S ₁ of each layer described above does not necessarily reflect the performance of the servers S ₃ , S ₂ , S ₁ . That is, it is not necessary for the servers in each layer to be aware of the processing time of servers lower than the own device in advance. In addition, as described above, in the configuration in which the completion time is calculated including the processing time in consideration of the processing load in the servers of each hierarchy, the completion time reflecting the variation in the processing time due to the variation in the processing load is given to the upper server. Reported and used to replace completion timer value.

  Therefore, timeout control is properly performed based on the completion timer value replaced in this way.

For example, when at least one of the processing times in the servers S ₃ , S ₂ , S ₁ of each layer becomes larger than the default value, the timing at which a response is returned from the lower server is the original completion as shown in FIG. The timeout control based on the timer value is eliminated. However, as described above, by replacing the original completion timer value with the new completion timer value, the server S ₄ can receive a response normally returned from the lower server, and the response is unfairly excluded. You can avoid that.

Further, as described above, by reflecting the performance of the servers S ₃ , S ₂ , S ₁ of each layer and the influence of the processing load, it is possible to set a new completion timer value that is not excessive or insufficient. Therefore, it is possible to prevent unnecessary waiting time from occurring when an abnormality occurs in any of the servers.

Similarly, also in the servers S ₃ and S ₂ located in the intermediate hierarchy, the completion timer value is replaced according to the completion time report from the server located in the lower hierarchy than the respective servers, and the completion after the replacement is completed. Timeout control based on the timer value is performed.

In the example shown in FIG. 6, when the server S ₂ receives the completion time report (2-1) including the completion time obtained by the completion time calculation process executed when the processing request REQ2 is received. , by the server _{S 3,} perform a completion time calculation process _{P 32,} it is sending completion time reported (3-2).

  As described above, the timer value management system can be configured so that the completion time calculation process is performed at all the processing request reception timing and the completion time report reception timing, and the calculated completion time is reported. It is also possible to omit the reporting of the intermediate completion time including the default value and report only the final completion time.

  Further, after the completion time information is collected, only when the completion time different from the already collected completion timer value is reported from the lower server, the completion time including the processing time of the own device is calculated and the upper server is calculated. By reporting to, the communication load for the timer value management process can be reduced.

Although omitted in the timer value management system shown in FIG. 1, the upper server 210 may have a configuration corresponding to the completion time calculation unit 224 and the completion time report unit 226 provided in the intermediate server 220. . Similarly, the lower server 230 can also have a configuration corresponding to the completion time collection unit 225.
(Embodiment 2)
FIG. 7 shows another embodiment of the timer value management system.

  7 that are the same as those shown in FIG. 1 are given the same reference numerals, and descriptions thereof are omitted.

In the server system shown in FIG. 7, the processing request sent from the upper server 210 is relayed to at least one of the two lower servers S ₁₀ and S ₂₀ by the intermediate server 220.

  The upper server 210 and the intermediate server 220 illustrated in FIG. 7 include polling control units 216 and 228 in place of the timeout control units 212 and 222. Further, the completion timer value tables 213 and 223 shown in FIG. 7 hold the minimum value and the maximum value of the completion time corresponding to the combination of the request type and the request destination server ID.

Further, the lower servers S ₁₀ and S ₂₀ shown in FIG. 7 are configured in the same manner as the lower server 230 shown in FIG.

FIG. 8 shows an example of the completion timer value table. In the example shown in FIG. 8, the completion indicating the completion time reported from the lower server corresponding to the combination of the request types A, B, and C, the server ID (S ₁₀ ), and the server ID (S ₂₀ ). Stores the minimum and maximum timer values. FIG. 9 is a sequence diagram for explaining the operation for managing the completion timer value.

The completion time calculation unit 234 of the lower servers S ₁₀ and S ₂₀ obtains the minimum value and the maximum value of the completion time until the processing of the own device is completed for the processing specified by the processing request (reference symbol P _{S1 in} FIG. 9). , P _s2 ). The calculated minimum value and maximum value are passed to the report collection unit 226 of the intermediate server 220 via the completion time report unit 233 (completion time reports (S-1) and (S-2 shown in FIG. 9)). )), And is used for updating the completion timer value table 223.

In this way, based on the completion timer value table 223 that reflects the range of completion times of the lower servers S ₁₀ and S ₂₀ , the polling control unit 228 waits after the change corresponding to the minimum value of each completion time. Polling processing is started after elapse of time t ₁ and t ₂ . The polling process is repeated over the polling period from the minimum value to the maximum value of the completion time described above until a processing result is obtained from the lower servers S10 and S20.

On the other hand, in response to the reception of the completion time report described above, the completion time calculation unit 229 of the intermediate server 220 is based on the received completion time range and the minimum and maximum values of the time required for processing of the own device. determining the range of the completion time for processing up to the own apparatus (reference numeral P _M2 of FIG. 9). Then, the obtained completion time range is passed to the completion report collection unit 214 of the host device via the completion time report unit 226 (completion time report (M-2) in FIG. 9), and the completion timer value table 213 is received. It is used for the update process.

  In this way, the completion time range up to the intermediate server 220 is reflected in the completion timer value table 213. Based on this completion timer value table 213, the polling control unit 216 of the upper server 210 determines whether or not the intermediate server 220 has received the waiting time after the change corresponding to the minimum value of the completion time reported from the intermediate server 220. Start polling processing. The polling process is repeated over the polling period from the minimum value to the maximum value of the completion time described above until a processing result is obtained from the intermediate server 220.

For example, the like if the child server S ₂₀ is replaced by a fast server processing speed, by time ready collected as described above is reflected, the new completion timer value, and replaces the default value It becomes shorter than the original value corresponding to the processing speed of the previous server device. Thus, it has improved the performance of the subordinate server S ₂₀ by the replacement, the intermediate server 220 or host server 210 is reflected in the time to wait for the start of the polling process. As a result, generation of useless waiting time can be prevented and the processing speed can be increased.

Regarding the above description, the following items are further disclosed.
(Supplementary note 1) A processing request provided to each server included in a multi-tier server system having three or more layers and sent to a server lower than the own device is located in a layer up to the destination server. Information collecting means for collecting completion time information indicating a time until processing by the server is completed;
A completion timer value table provided in each of the servers, corresponding to the transmitted processing request, holding a completion timer value based on the collected completion time information, and providing time-out control for the processing request;
Completion time report including completion time information indicating the completion time, which is provided in each server, obtains a completion time including the processing time of the own device based on the collected completion time information and the time required for processing in the own device. And a reporting means for sending the message to a server located in a higher hierarchy than the own device. (1)
(Appendix 2) The reporting means
When the completion time indicated by the completion time information collected by the collection means is different from the completion timer value held in the completion timer value table, completion time information indicating a new completion time is sent to the upper server. The timer value management system according to supplementary note 1, further comprising change determination means for use in report processing. (2)
(Appendix 3) The reporting means is
Considering the load state of the own device, the load reflecting means for obtaining the processing time by the own device for the process specified in the processing request,
Supplementary note 1 characterized in that the time required for processing up to the device obtained based on the processing time in which the load is reflected and the collected required time is used for reporting to a server in a higher hierarchy. Timer value management system.
(Appendix 4) The reporting means
When processing requests are relayed in parallel to a plurality of servers located in a lower hierarchy than the own device, the maximum required time is detected from the required times included in the collected required time information, and the maximum required time is detected. The timer value management system according to appendix 1, further comprising a parallel reflection unit for providing time to a report to an upper layer server.
(Appendix 5) The reporting means
When processing requests are relayed serially to a plurality of servers located in a lower hierarchy than the own device, the required time included in the collected required time information is added to calculate the total required time, and this required time The timer value management system according to claim 1, further comprising: serial reflection means for providing a sum of the above to a server of a higher hierarchy.
(Supplementary Note 6) A processing request that is provided in each server included in a multi-stage server system having three or more hierarchies and is sent to a server lower than its own device is located in the hierarchy up to the destination server. An information collecting means for collecting completion time information indicating a minimum value and a maximum value of a time until processing by the server is completed;
A completion timer value table provided in each of the servers, holding a completion timer value based on the collected completion time information corresponding to the transmitted processing request, and providing polling control for the processing request;
Provided in each of the servers, a minimum value and a maximum value of the completion time including the processing time of the own device are obtained based on the collected completion time information and the time required for processing in the own device, and the range of the completion time is indicated. A timer value management system comprising: reporting means for sending a completion time report including completion time information to a server located in a higher hierarchy than the own device. (3)
(Supplementary Note 7) Any one of server devices included in a multistage server system having three or more layers,
Information collection means for collecting completion time information indicating a time until processing by a server located in a hierarchy up to the transmission destination server is completed for a processing request transmitted to a server lower than the own device;
Corresponding to the sent processing request, a completion timer value is held based on the collected completion time information and used for timeout control for the processing request;
Based on the collected completion time information and the time required for processing in the own device, a completion time including the processing time of the own device is obtained, and a completion time report including the completion time information indicating the completion time is higher than the own device. And a reporting means for sending to a server located in the hierarchy of the server. (4)
(Supplementary Note 8) Any one of server devices included in a multi-tier server system having three or more layers,
Information collection for collecting completion time information indicating a minimum value and a maximum value of a time required for completion of processing by a server located in a hierarchy up to the destination server with respect to a processing request transmitted to a server lower than the own device Means,
A completion timer value table provided in each of the servers, corresponding to the transmitted processing request, holding a completion timer value based on the collected completion time information, and providing polling control for the processing request;
Provided in each of the servers, a minimum value and a maximum value of the completion time including the processing time of the own device are obtained based on the collected completion time information and the time required for processing in the own device, and the range of the completion time is indicated. A server device comprising: reporting means for transmitting a completion time report including completion time information to a server located in a higher hierarchy than the own device. (5)

210 Host servers 211, 221, 231 Processing control units 212, 222 Timeout control units 213, 223 Completion timer value tables 214, 225 Report collection units 215, 227 Server information tables 216, 228 Polling control unit 220 Intermediate servers 224, 232, 229 , 234 Completion time calculation unit 226, 233 Completion time report unit 241 Processing time table 243 Load information collection unit 244 Processing time calculation unit 245 Completion time calculation unit 246 Calculation control unit

Claims (5)

A processing request provided to each server included in a multistage server system having three or more layers and sent to a server lower than the own device is processed by a server located in the hierarchy up to the destination server. Information collection means for collecting completion time information indicating the time until completion;
A completion timer value table provided in each of the servers, corresponding to the transmitted processing request, holding a completion timer value based on the collected completion time information, and providing time-out control for the processing request;
Completion time report including completion time information indicating the completion time, which is provided in each server, obtains a completion time including the processing time of the own device based on the collected completion time information and the time required for processing in the own device. And a reporting means for sending the message to a server located in a higher hierarchy than the own device. In the timer value management system according to claim 1,
The reporting means includes
When the completion time indicated by the completion time information collected by the collection means is different from the completion timer value held in the completion timer value table, completion time information indicating a new completion time is sent to the upper server. A timer value management system comprising a change judging means for reporting processing. A processing request provided to each server included in a multistage server system having three or more layers and sent to a server lower than the own device is processed by a server located in the hierarchy up to the destination server. An information collecting means for collecting completion time information indicating a minimum value and a maximum value of time until completion;
A completion timer value table provided in each of the servers, holding a completion timer value based on the collected completion time information corresponding to the transmitted processing request, and providing polling control for the processing request;
Provided in each of the servers, a minimum value and a maximum value of the completion time including the processing time of the own device are obtained based on the collected completion time information and the time required for processing in the own device, and the range of the completion time is indicated. A timer value management system comprising: reporting means for sending a completion time report including completion time information to a server located at a higher hierarchy than the own device. Any of server devices included in a multi-tier server system having three or more layers,
Information collection means for collecting completion time information indicating a time until processing by a server located in a hierarchy up to the transmission destination server is completed for a processing request transmitted to a server lower than the own device;
Corresponding to the sent processing request, a completion timer value is held based on the collected completion time information and used for timeout control for the processing request;
Based on the collected completion time information and the time required for processing in the own device, a completion time including the processing time of the own device is obtained, and a completion time report including the completion time information indicating the completion time is higher than the own device. And a reporting means for sending to a server located in the hierarchy of the server. Any of server devices included in a multi-tier server system having three or more layers,
Information collection for collecting completion time information indicating a minimum value and a maximum value of a time required for completion of processing by a server located in a hierarchy up to the destination server with respect to a processing request sent to a server lower than the own device Means,
A completion timer value table provided in each of the servers, holding a completion timer value based on the collected completion time information corresponding to the transmitted processing request, and providing polling control for the processing request;
Provided in each of the servers, a minimum value and a maximum value of the completion time including the processing time of the own device are obtained based on the collected completion time information and the time required for processing in the own device, and the range of the completion time is indicated. A server device comprising: reporting means for transmitting a completion time report including completion time information to a server located in a higher hierarchy than the own device.

Images