TWI785644B - Dynamic management system and management host - Google Patents

Dynamic management system and management host Download PDF

Info

Publication number
TWI785644B
TWI785644B TW110121375A TW110121375A TWI785644B TW I785644 B TWI785644 B TW I785644B TW 110121375 A TW110121375 A TW 110121375A TW 110121375 A TW110121375 A TW 110121375A TW I785644 B TWI785644 B TW I785644B
Authority
TW
Taiwan
Prior art keywords
files
server
host
processing module
processed
Prior art date
Application number
TW110121375A
Other languages
Chinese (zh)
Other versions
TW202248851A (en
Inventor
譚蓓華
方鈞鶴
楊明憲
Original Assignee
中國信託商業銀行股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中國信託商業銀行股份有限公司 filed Critical 中國信託商業銀行股份有限公司
Priority to TW110121375A priority Critical patent/TWI785644B/en
Application granted granted Critical
Publication of TWI785644B publication Critical patent/TWI785644B/en
Publication of TW202248851A publication Critical patent/TW202248851A/en

Links

Images

Landscapes

  • Preparation Of Compounds By Using Micro-Organisms (AREA)
  • Computer And Data Communications (AREA)
  • Multi Processors (AREA)

Abstract

一種動態管理系統包含多個伺服主機及一管理主機,該管理主機包括一處理模組及一通訊模組,該通訊模組電連接該處理模組,並能夠與該等伺服主機建立連線。該處理模組經由該通訊模組接收待處理的多個工作檔案,並先將該等工作檔案之其中部分者分配給該等伺服主機作處理,以計算出每一該伺服主機的一運算能力,再根據每一該伺服主機的該運算能力及剩餘待處理的該等工作檔案的數量,將剩餘待處理的該等工作檔案分配給該等伺服主機作處理,以實現一種根據每一該伺服主機的該運算能力分配該等工作檔案的動態管理系統。A dynamic management system includes multiple server hosts and a management host. The management host includes a processing module and a communication module. The communication module is electrically connected to the processing module and can establish connections with the server hosts. The processing module receives a plurality of work files to be processed through the communication module, and first distributes some of the work files to the server hosts for processing, so as to calculate a computing power of each server host , and then according to the computing power of each server host and the number of the remaining work files to be processed, the remaining work files to be processed are allocated to the server hosts for processing, so as to realize a process based on each server The computing power of the host computer is allocated to the dynamic management system of the work files.

Description

動態管理系統及管理主機Dynamic management system and management host

本發明是有關於一種管理系統及管理主機,特別是指一種有效管理伺服器資源的動態管理系統及管理主機。The present invention relates to a management system and a management host, in particular to a dynamic management system and a management host for effectively managing server resources.

現有的金融保險業者利用電腦處理大量案件時,最普遍的作法是採用多個伺服主機,並以其中一者作為一主控裝置(Master),藉由該主控裝置監看其餘該等伺服主機的運作狀況,以獲知哪些伺服主機處於閒置狀態,進而將所接收到的所有案件,逐一分配給處於閒置狀態的每一該伺服主機,如此持續執行監看與分配工作的運作直到所有案件處理完成,而能夠實現一種分工處理的管理系統。然而,是否存有其他更具效率以分配案件工作的管理系統便成為一個待解決的問題。When existing financial and insurance companies use computers to handle a large number of cases, the most common method is to use multiple server hosts, and use one of them as a master control device (Master), and use the master control device to monitor the remaining server hosts In order to know which server hosts are idle, and then distribute all received cases to each idle server one by one, so that the monitoring and distribution work continues until all cases are processed , and can realize a management system of division of labor. However, it remains an open question whether there are other more efficient management systems for allocating casework.

因此,本發明的目的,即在提供一種有效管理伺服器資源的動態管理系統及管理主機。Therefore, the object of the present invention is to provide a dynamic management system and a management host for effectively managing server resources.

於是,本發明之一觀點,提供一種動態管理系統,包含多個伺服主機及一管理主機。該管理主機包括一處理模組及一通訊模組,該通訊模組電連接該處理模組,並能夠與該等伺服主機建立連線。Therefore, an aspect of the present invention provides a dynamic management system, which includes a plurality of server hosts and a management host. The management host includes a processing module and a communication module. The communication module is electrically connected to the processing module and can establish connections with the server hosts.

其中,該處理模組經由該通訊模組接收待處理的多個工作檔案(Task),並先將該等工作檔案之其中部分者經由該通訊模組分配給該等伺服主機作處理,以計算出每一該伺服主機的一運算能力,再根據每一該伺服主機的該運算能力及剩餘待處理的該等工作檔案的數量,將剩餘待處理的該等工作檔案經由該通訊模組分配給該等伺服主機作處理。Wherein, the processing module receives a plurality of work files (Task) to be processed through the communication module, and first distributes some of the work files to the server hosts for processing through the communication module, so as to calculate According to the computing power of each server host and the number of remaining work files to be processed, distribute the remaining work files to be processed via the communication module These server hosts do the processing.

在一些實施態樣中,其中,該管理主機的該處理模組將所接收且待處理的該等工作檔案依所接收的時間順序排列,並將待處理的該等工作檔案依序分配給該等伺服主機之其中尚未計算出該運算能力者,以獲知每一該伺服主機處理完成單一個該工作檔案所需要的一處理時間,進而計算出每一該伺服主機的該運算能力等於對應的該處理時間。In some implementation aspects, wherein, the processing module of the management host arranges the received and to-be-processed work files according to the received time order, and sequentially assigns the to-be-processed work files to the Wait for one of the server hosts that has not yet calculated the computing power to know the processing time required for each server host to complete a single job file, and then calculate the computing power of each server host equal to the corresponding processing time.

在一些實施態樣中,其中,該管理主機的該處理模組根據一第一公式及一第二公式,並以拉格朗日乘數法(Lagrange multiplier)求出要分配給每一該伺服主機的剩餘待處理的該等工作檔案的一檔案數量。該第一公式是:min(該等伺服主機所分別分配的該等檔案數量分別乘以對應的該等運算能力之和)。該第二公式是:該等伺服主機所分別分配的該等檔案數量之和=剩餘待處理的該等工作檔案之數量總和。該處理模組再根據每一該伺服主機所對應的該檔案數量,將剩餘待處理的該等工作檔案經由該通訊模組分配給該等伺服主機作處理。In some implementation aspects, wherein, the processing module of the management host obtains the information to be allocated to each server according to a first formula and a second formula, and uses Lagrange multiplier method (Lagrange multiplier). A file quantity of the remaining work files of the host to be processed. The first formula is: min (the number of files allocated by the server hosts multiplied by the sum of the corresponding computing capabilities). The second formula is: the sum of the number of files allocated by the server hosts = the sum of the number of remaining work files to be processed. The processing module distributes the remaining work files to be processed to the server hosts through the communication module according to the number of files corresponding to each server host.

在一些實施態樣中,其中,該處理模組執行步驟S1及S2。In some implementation aspects, the processing module executes steps S1 and S2.

於步驟S1,當該處理模組判斷該等伺服主機的該等檔案數量之其中至少一者不等於整數時,將其中一個該檔案數量作四捨五入計算而獲得整數化的該檔案數量。In step S1, when the processing module judges that at least one of the file quantities of the server hosts is not equal to an integer, one of the file quantities is rounded up to obtain the integerized file quantity.

於步驟S2,該處理模組將整數化的該檔案數量代入該第一公式及該第二公式後再以拉格朗日乘數法求出要分配給其餘的每一該伺服主機的剩餘待處理的該等工作檔案的該檔案數量。In step S2, the processing module substitutes the integerized number of files into the first formula and the second formula, and then uses the Lagrangian multiplier method to obtain the remaining waiting time allocated to each of the remaining server hosts. the number of such job files processed.

該處理模組重覆執行步驟S1及S2,直到每一該伺服主機所對應的該檔案數量都等於整數為止,該處理模組才將剩餘待處理且等於每一該伺服主機所對應的該檔案數量的該等工作檔案,經由該通訊模組分配給對應的該伺服主機作處理。The processing module repeatedly executes steps S1 and S2 until the number of files corresponding to each of the server hosts is equal to an integer, and the processing module will remain to be processed and equal to the number of files corresponding to each of the server hosts The number of these work files are allocated to the corresponding server host for processing through the communication module.

於是,本發明之一觀點,提供一種管理主機,適用於多個伺服主機,並包含一處理模組及電連接該處理模組的一通訊模組。其中,該處理模組經由該通訊模組接收待處理的多個工作檔案,並先將該等工作檔案之其中部分者經由該通訊模組分配給該等伺服主機作處理,以計算出每一該伺服主機的一運算能力,再根據每一該伺服主機的該運算能力及剩餘待處理的該等工作檔案的數量,將剩餘待處理的該等工作檔案經由該通訊模組分配給該等伺服主機作處理。Therefore, an aspect of the present invention provides a management host, which is suitable for multiple servo hosts, and includes a processing module and a communication module electrically connected to the processing module. Wherein, the processing module receives a plurality of work files to be processed through the communication module, and first distributes some of the work files to the server hosts through the communication module for processing, so as to calculate each According to the computing capability of the server host, and according to the computing capability of each server host and the number of the remaining work files to be processed, the remaining work files to be processed are distributed to the servers through the communication module Host for processing.

在一些實施態樣中,其中,該管理主機的該處理模組將所接收且待處理的該等工作檔案依所接收的時間順序排列,並將待處理的該等工作檔案依序分配給該等伺服主機之其中尚未計算出該運算能力者,以獲知每一該伺服主機處理完成單一個該工作檔案所需要的一處理時間,進而計算出每一該伺服主機的該運算能力等於對應的該處理時間。In some implementation aspects, wherein, the processing module of the management host arranges the received and to-be-processed work files according to the received time order, and sequentially assigns the to-be-processed work files to the Wait for one of the server hosts that has not yet calculated the computing power to know the processing time required for each server host to complete a single job file, and then calculate the computing power of each server host equal to the corresponding processing time.

在一些實施態樣中,其中,該管理主機的該處理模組根據一第一公式及一第二公式,並以拉格朗日乘數法求出要分配給每一該伺服主機的剩餘待處理的該等工作檔案的一檔案數量。該第一公式是:min(該等伺服主機所分別分配的該等檔案數量分別乘以對應的該等運算能力之和)。該第二公式是:該等伺服主機所分別分配的該等檔案數量之和=剩餘待處理的該等工作檔案之數量總和。該處理模組再根據每一該伺服主機所對應的該檔案數量,將剩餘待處理的該等工作檔案經由該通訊模組分配給該等伺服主機作處理。In some implementation aspects, wherein, the processing module of the management host calculates the remaining waiting time allocated to each of the server hosts by using a Lagrangian multiplier method according to a first formula and a second formula. A file quantity of the job files processed. The first formula is: min (the number of files allocated by the server hosts multiplied by the sum of the corresponding computing capabilities). The second formula is: the sum of the number of files allocated by the server hosts = the sum of the number of remaining work files to be processed. The processing module distributes the remaining work files to be processed to the server hosts through the communication module according to the number of files corresponding to each server host.

在一些實施態樣中,其中,該處理模組執行步驟S1及S2。於步驟S1,當該處理模組判斷該等伺服主機的該等檔案數量之其中至少一者不等於整數時,將其中一個該檔案數量作四捨五入計算而獲得整數化的該檔案數量。In some implementation aspects, the processing module executes steps S1 and S2. In step S1, when the processing module judges that at least one of the file quantities of the server hosts is not equal to an integer, one of the file quantities is rounded up to obtain the integerized file quantity.

於步驟S2,該處理模組將整數化的該檔案數量代入該第一公式及該第二公式後再以拉格朗日乘數法求出要分配給其餘的每一該伺服主機的剩餘待處理的該等工作檔案的該檔案數量。In step S2, the processing module substitutes the integerized number of files into the first formula and the second formula, and then uses the Lagrangian multiplier method to obtain the remaining waiting time allocated to each of the remaining server hosts. the number of such job files processed.

該處理模組重覆執行步驟S1及S2,直到每一該伺服主機所對應的該檔案數量都等於整數為止,該處理模組才將剩餘待處理且等於每一該伺服主機所對應的該檔案數量的該等工作檔案,經由該通訊模組分配給對應的該伺服主機作處理。The processing module repeatedly executes steps S1 and S2 until the number of files corresponding to each of the server hosts is equal to an integer, and the processing module will remain to be processed and equal to the number of files corresponding to each of the server hosts The number of these work files are allocated to the corresponding server host for processing through the communication module.

本發明的功效在於:藉由該管理主機的該處理模組先計算出每一該伺服主機的該運算能力,再根據每一該伺服主機的該運算能力分配剩餘待處理的該等工作檔案,而能實現一種有效管理伺服器資源的動態管理系統及管理主機。The effect of the present invention lies in that: the processing module of the management host firstly calculates the computing capability of each of the server hosts, and then allocates the remaining work files to be processed according to the computing capability of each of the server hosts, Therefore, a dynamic management system and a management host for effectively managing server resources can be realized.

在本發明被詳細描述之前,應當注意在以下的說明內容中,類似的元件是以相同的編號來表示。Before the present invention is described in detail, it should be noted that in the following description, similar elements are denoted by the same numerals.

參閱圖1,本發明動態管理系統之一實施例,包含多個伺服主機91~93及一管理主機1。該管理主機1包括一處理模組11及一通訊模組12,該通訊模組12電連接該處理模組11,並能夠與該等伺服主機91~93建立連線。在本實施例中,每一該伺服主機91~93及該管理主機1都屬於一銀行業者,並例如是一種伺服器設備,該處理模組11例如是一個或多個中央處理器(CPU),該通訊模組12例如是一網路卡、一網路晶片,或一無線網路模組等。Referring to FIG. 1 , an embodiment of the dynamic management system of the present invention includes a plurality of server hosts 91 to 93 and a management host 1 . The management host 1 includes a processing module 11 and a communication module 12. The communication module 12 is electrically connected to the processing module 11 and can establish connections with the server hosts 91-93. In this embodiment, each of the server hosts 91-93 and the management host 1 belongs to a banker, and is, for example, a server device, and the processing module 11 is, for example, one or more central processing units (CPUs). , the communication module 12 is, for example, a network card, a network chip, or a wireless network module.

該管理主機1的該處理模組11經由該通訊模組12接收待處理的多個工作檔案(Task),舉例來說,該通訊模組12是接收來自其他伺服主機或電腦設備的該等工作檔案,每一該工作檔案例如是該銀行業者所要處理的一紓困案的電子檔案。The processing module 11 of the management host 1 receives a plurality of task files (Task) to be processed through the communication module 12. For example, the communication module 12 receives the tasks from other server hosts or computer equipment files, each of the working files is, for example, an electronic file of a bailout case to be handled by the banker.

該管理主機1的該處理模組11將所接收且待處理的該等工作檔案依所接收的時間順序排列,並將待處理的該等工作檔案依序分配給該等伺服主機91~93之其中尚未計算出該運算能力者,以獲知每一該伺服主機91~93處理完成單一個該工作檔案所需要的一處理時間,進而計算出每一該伺服主機91~93的一運算能力等於對應的該處理時間。The processing module 11 of the management host 1 arranges the work files received and to be processed according to the received time order, and distributes the work files to be processed to the server hosts 91~93 in sequence Among them, those who have not yet calculated the computing power, so as to know the processing time required for each of the server hosts 91-93 to process a single work file, and then calculate the computing power of each of the servo hosts 91-93 to be equal to the corresponding of the processing time.

再參閱圖2,舉例來說,該等伺服主機#1~#5的數量是五個,該管理主機1依序所接收到的該等工作檔案W1~W17的數量是17個。在時間=0時,該處理模組11完全不知道每一該伺服主機#1~#5的該運算能力,因此,該處理模組11將該等工作檔案W1~W5依序分配給該等伺服主機#1~#5。在時間=1時,該處理模組11獲知該伺服主機#1完成該工作檔案W1,接著,依序將該等工作檔案W6~W8再次分配給該伺服主機#1。直到時間=4時,該等伺服主機#2~#5都完成該等工作檔案W2~W5。此時,該處理模組11計算而獲得該等伺服主機#1~#5的該等運算能力分別是1、4、4、4、及4。要特別強調的是時間=1是指一個單位時間,例如是5秒或其他時間。Referring to FIG. 2 again, for example, the number of the server hosts #1~#5 is five, and the number of the work files W1~W17 received by the management host 1 is 17. When time=0, the processing module 11 is completely unaware of the computing capability of each of the servo hosts #1~#5, therefore, the processing module 11 sequentially distributes the work files W1~W5 to the Servo host #1~#5. When time=1, the processing module 11 knows that the server host #1 has completed the work file W1, and then sequentially distributes the work files W6-W8 to the server host #1 again. Until time=4, the server hosts #2~#5 have completed the work files W2~W5. At this time, the processing module 11 calculates and obtains that the computing capabilities of the server hosts #1-#5 are 1, 4, 4, 4, and 4, respectively. It should be emphasized that time=1 refers to a unit time, for example, 5 seconds or other time.

該管理主機1的該處理模組11根據一第一公式及一第二公式,並以拉格朗日乘數法(Lagrange multiplier)求出要分配給每一該伺服主機#1~#5的剩餘待處理的該等工作檔案的一檔案數量。According to a first formula and a second formula, the processing module 11 of the management host 1 calculates the value to be assigned to each of the server hosts #1~#5 by using the Lagrange multiplier method. A file quantity of the work files remaining to be processed.

該第一公式是:min(該等伺服主機#1~#5所分別分配的該等檔案數量X1~X5分別乘以對應的該等運算能力之和),min()表示求最小值。該第二公式是:該等伺服主機#1~#5所分別分配的該等檔案數量X1~X5之和=剩餘待處理的該等工作檔案之數量總和。承續前例,舉例來說,該第一公式是min(X1*1+X2*4+X3*4+ X4*4+X5*4),該第二公式是X1+X2+X3+X4+X5=17-8=9。則以拉格朗日乘數法所求出的該等檔案數量X1~X5分別是4.5、1.125、1.125、1.125、1.125。The first formula is: min (multiply the number of files X1~X5 respectively allocated by the server hosts #1~#5 by the sum of the corresponding computing capabilities), and min() means finding the minimum value. The second formula is: the sum of the numbers X1-X5 of the files allocated by the server hosts #1-#5=the sum of the number of remaining work files to be processed. Continuing from the previous example, for example, the first formula is min(X1*1+X2*4+X3*4+ X4*4+X5*4), and the second formula is X1+X2+X3+X4+X5 =17-8=9. Then the numbers X1~X5 of the files calculated by the Lagrange multiplier method are 4.5, 1.125, 1.125, 1.125, and 1.125 respectively.

該處理模組11再根據每一該伺服主機#1~#5所對應的該檔案數量X1~X5,將剩餘待處理的該等工作檔案經由該通訊模組12分配給該等伺服主機#1~#5作處理。更詳細地說,該處理模組11執行步驟S1及S2。The processing module 11 distributes the remaining work files to be processed to the server hosts #1 through the communication module 12 according to the number of files X1~X5 corresponding to each of the server hosts #1~#5 ~#5 for processing. More specifically, the processing module 11 executes steps S1 and S2.

於步驟S1,當該處理模組11判斷該等伺服主機#1~#5的該等檔案數量X1~X5之其中至少一者不等於整數時,將其中一個該檔案數量X1~X5作四捨五入計算而獲得整數化的該檔案數量X1~X5。承續前例,舉例來說,該處理模組11將對應該伺服主機#5的該檔案數量X5(即1.125)作四捨五入而獲得整數化的該檔案數量X5(即1)。In step S1, when the processing module 11 judges that at least one of the file numbers X1~X5 of the server hosts #1~#5 is not equal to an integer, one of the file numbers X1~X5 is rounded up for calculation And obtain the integerized numbers X1˜X5 of the files. Continuing from the previous example, for example, the processing module 11 rounds up the file number X5 (ie 1.125) corresponding to the server host #5 to obtain the integerized file number X5 (ie 1).

於步驟S2,該處理模組11將整數化的該檔案數量X5代入該第一公式及該第二公式後再以拉格朗日乘數法求出要分配給其餘的每一該伺服主機#1~#5的剩餘待處理的該等工作檔案的該檔案數量X1~X4。承續前例,舉例來說,更新後的該第一公式是min(X1*1+X2*4+X3*4+X4*4+1*4),更新後的該第二公式是X1+X2+X3+X4+1=9。則以拉格朗日乘數法所求出的該等檔案數量X1~X5分別是4.571、1.143、1.143、1.143、及1。In step S2, the processing module 11 substitutes the integerized file quantity X5 into the first formula and the second formula, and then uses the Lagrangian multiplier method to obtain # to be distributed to each of the remaining server hosts. Numbers X1~X4 of the remaining work files 1~#5 to be processed. Continuing from the previous example, for example, the updated first formula is min(X1*1+X2*4+X3*4+X4*4+1*4), and the updated second formula is X1+X2 +X3+X4+1=9. Then the numbers X1~X5 of the files calculated by the Lagrangian multiplier method are 4.571, 1.143, 1.143, 1.143, and 1, respectively.

接著,該處理模組11重覆執行步驟S1及S2,直到每一該伺服主機#1~#5所對應的該檔案數量X1~X5都等於整數為止。承續前例,舉例來說,再次執行步驟S1將該伺服主機#4的該檔案數量X4(即1.143)作四捨五入而獲得整數化的該檔案數量X4(即1)。再次執行步驟S2獲得更新後的該第一公式是min(X1*1+X2*4+X3*4+1*4+1*4),更新後的該第二公式是X1+X2+X3+1+1=9。則以拉格朗日乘數法所求出的該等檔案數量X1~X5分別是4.67、1.17、1.17、1、及1。Next, the processing module 11 repeatedly executes steps S1 and S2 until the number of files X1-X5 corresponding to each of the server hosts #1-#5 is equal to an integer. Continuing from the previous example, for example, execute step S1 again to round up the file number X4 (ie 1.143) of the server host #4 to obtain the integerized file number X4 (ie 1). Execute step S2 again to obtain the updated first formula is min(X1*1+X2*4+X3*4+1*4+1*4), and the updated second formula is X1+X2+X3+ 1+1=9. Then the numbers X1~X5 of the files calculated by the Lagrange multiplier method are 4.67, 1.17, 1.17, 1, and 1, respectively.

接著,再次執行步驟S1將該伺服主機#3的該檔案數量X3(即1.17)作四捨五入而獲得整數化的該檔案數量X3(即1)。再次執行步驟S2獲得更新後的該第一公式是min(X1*1+X2*4+1*4+ 1*4+1*4),更新後的該第二公式是X1+X2+1+1+1=9。則以拉格朗日乘數法所求出的該等檔案數量X1~X5分別是4.8、1.2、1、1、及1。Next, execute step S1 again to round up the file quantity X3 (ie 1.17) of the server host #3 to obtain the integerized file quantity X3 (ie 1). Execute step S2 again to obtain the updated first formula is min(X1*1+X2*4+1*4+ 1*4+1*4), and the updated second formula is X1+X2+1+ 1+1=9. Then the numbers X1~X5 of the files calculated by the Lagrange multiplier method are 4.8, 1.2, 1, 1, and 1, respectively.

接著,再次執行步驟S1將該伺服主機#2的該檔案數量X2(即1.2)作四捨五入而獲得整數化的該檔案數量X2(即1)。再次執行步驟S2獲得更新後的該第一公式是min(X1*1+1*4+ 1*4+1*4+1*4),更新後的該第二公式是X1+1+1+1+1=9。則以拉格朗日乘數法所求出的該等檔案數量X1~X5分別是5、1、1、1、及1。Next, execute step S1 again to round up the file quantity X2 (ie 1.2) of the server host #2 to obtain the integerized file quantity X2 (ie 1). Execute step S2 again to obtain the updated first formula is min(X1*1+1*4+1*4+1*4+1*4), and the updated second formula is X1+1+1+ 1+1=9. Then the numbers X1~X5 of the files calculated by the Lagrange multiplier method are 5, 1, 1, 1, and 1 respectively.

該處理模組11才將剩餘待處理且等於每一該伺服主機#1~#5所對應的該檔案數量X1~X5的該等工作檔案,經由該通訊模組12分配給對應的該伺服主機#1~#5作處理。承續前例,舉例來說,該處理模組11根據該等伺服主機#1~#5的該等檔案數量X1~X5(即5、1、1、1、及1),將該等工作檔案W9~W13(即5個工作檔案)分配給該伺服主機#1,且將該等工作檔案W14~W17分別分配給該等伺服主機#2~#5(即各分配1個工作檔案)。The processing module 11 distributes the remaining work files to be processed and equal to the number of files X1~X5 corresponding to each of the server hosts #1~#5 to the corresponding server hosts via the communication module 12 #1~#5 for processing. Continuing from the previous example, for example, the processing module 11 converts the working files to W9~W13 (ie, 5 work files) are assigned to the server host #1, and the work files W14~W17 are respectively assigned to the server hosts #2~#5 (ie, 1 work file is assigned to each).

另外要特別補充說明的是:在本實施例中,每一該工作檔案所需要的運算量都相同,而在其他的實施例中,每一該工作檔案所需要的運算量也可以是不相同。此外,每一該伺服主機#1~#5完成第一個該工作檔案的時間點也可以不相同,該管理主機1的該處理模組11能夠根據已經計算出的該至少一運算能力,動態更新該第一工作及該第二公式,以在不同的時間點隨時動態且即時地計算出如何分配剩餘的該等工作檔案給哪些該等伺服主機#1~#5。In addition, it should be specially supplemented that: in this embodiment, the amount of computation required by each work file is the same, but in other embodiments, the amount of computation required by each work file may also be different . In addition, the time point when each of the server hosts #1~#5 completes the first work file can also be different, and the processing module 11 of the management host 1 can dynamically calculate the at least one computing capability according to the calculated The first job and the second formula are updated to dynamically and real-time calculate how to allocate the remaining work files to the server hosts #1~#5 at different time points.

綜上所述,藉由該管理主機1的該處理模組11先計算出每一該伺服主機91~93的該運算能力,再根據每一該伺服主機91~ 93的該運算能力分配剩餘待處理的該等工作檔案,進而能夠實現一種有效管理伺服器資源的動態管理系統及管理主機1,故確實能達成本發明的目的。In summary, the processing module 11 of the management host 1 first calculates the computing capability of each of the servo hosts 91-93, and then allocates the remaining waiting time according to the computing capabilities of each of the servo hosts 91-93. The processed work files can further realize a dynamic management system for effectively managing server resources and the management host 1, so the purpose of the present invention can indeed be achieved.

惟以上所述者,僅為本發明的實施例而已,當不能以此限定本發明實施的範圍,凡是依本發明申請專利範圍及專利說明書內容所作的簡單的等效變化與修飾,皆仍屬本發明專利涵蓋的範圍內。But the above-mentioned ones are only embodiments of the present invention, and should not limit the scope of the present invention. All simple equivalent changes and modifications made according to the patent scope of the present invention and the content of the patent specification are still within the scope of the present invention. Within the scope covered by the patent of the present invention.

1:管理主機 11:處理模組 12:通訊模組 91~93:伺服主機 #1~#5:伺服主機 W1~W17:工作檔案 1: Manage the host 11: Processing module 12: Communication module 91~93: Servo host #1~#5: Servo Host W1~W17: Work files

本發明的其他的特徵及功效,將於參照圖式的實施方式中清楚地呈現,其中: 圖1是一方塊圖,說明本發明動態管理系統的一實施例;及 圖2是一示意圖,示例性地說明該實施例的多個工作檔案的一種分配態樣。 Other features and effects of the present invention will be clearly presented in the implementation manner with reference to the drawings, wherein: Figure 1 is a block diagram illustrating an embodiment of the dynamic management system of the present invention; and FIG. 2 is a schematic diagram illustrating an example of an allocation of multiple work files in this embodiment.

1:管理主機 1: Manage the host

11:處理模組 11: Processing module

12:通訊模組 12: Communication module

91~93:伺服主機 91~93: Servo host

Claims (8)

一種動態管理系統,包含:多個伺服主機;及一管理主機,包括一處理模組及一通訊模組,該通訊模組電連接該處理模組,並能夠與該等伺服主機建立連線;其中,該處理模組經由該通訊模組接收待處理的多個工作檔案(Task),並先將該等工作檔案之其中部分者經由該通訊模組分配給該等伺服主機作處理,以計算出每一該伺服主機的一運算能力,再根據每一該伺服主機的該運算能力及剩餘待處理的該等工作檔案的數量,將剩餘待處理的該等工作檔案經由該通訊模組分配給該等伺服主機作處理。 A dynamic management system, including: a plurality of server hosts; and a management host, including a processing module and a communication module, the communication module is electrically connected to the processing module, and can establish a connection with the server hosts; Wherein, the processing module receives a plurality of work files (Task) to be processed through the communication module, and first distributes some of the work files to the server hosts for processing through the communication module, so as to calculate According to the computing power of each server host and the number of remaining work files to be processed, distribute the remaining work files to be processed via the communication module These server hosts do the processing. 如請求項1所述的動態管理系統,其中,該管理主機的該處理模組將所接收且待處理的該等工作檔案依所接收的時間順序排列,並將待處理的該等工作檔案依序分配給該等伺服主機之其中尚未計算出該運算能力者,以獲知每一該伺服主機處理完成單一個該工作檔案所需要的一處理時間,並決定每一該伺服主機的該運算能力等於對應的該處理時間。 The dynamic management system as described in claim 1, wherein, the processing module of the management host arranges the work files received and to be processed according to the received time order, and arranges the work files to be processed according to Allocated to one of the server hosts that has not yet calculated the computing power, so as to know the processing time required for each server host to complete a single job file, and determine that the computing power of each server host is equal to corresponding to the processing time. 如請求項2所述的動態管理系統,其中,該管理主機的該處理模組根據一第一公式及一第二公式,並以拉格朗日乘數法(Lagrange multiplier)求出要分配給每一該伺服主機的剩餘待處理的該等工作檔案的一檔案數量,該第一公 式是:min(該等伺服主機所分別分配的該等檔案數量分別乘以對應的該等運算能力之和),該第二公式是:該等伺服主機所分別分配的該等檔案數量之和=剩餘待處理的該等工作檔案之數量總和,該處理模組再根據每一該伺服主機所對應的該檔案數量,將剩餘待處理的該等工作檔案經由該通訊模組分配給該等伺服主機作處理。 The dynamic management system as described in claim 2, wherein, the processing module of the management host calculates the distribution to A file quantity of the remaining work files to be processed for each server host, the first public The formula is: min (the sum of the number of files allocated by the server hosts multiplied by the corresponding computing power), the second formula is: the sum of the number of files allocated by the server hosts = The sum of the number of the remaining work files to be processed, the processing module distributes the remaining work files to the servers through the communication module according to the number of files corresponding to each server host Host for processing. 如請求項3所述的動態管理系統,其中,該處理模組執行步驟S1及S2,於步驟S1,當該處理模組判斷該等伺服主機的該等檔案數量之其中至少一者不等於整數時,將其中一個該檔案數量作四捨五入計算而獲得整數化的該檔案數量,於步驟S2,該處理模組將整數化的該檔案數量代入該第一公式及該第二公式後再以拉格朗日乘數法求出要分配給其餘的每一該伺服主機的剩餘待處理的該等工作檔案的該檔案數量,該處理模組重覆執行步驟S1及S2,直到每一該伺服主機所對應的該檔案數量都等於整數為止,該處理模組才將剩餘待處理且等於每一該伺服主機所對應的該檔案數量的該等工作檔案,經由該通訊模組分配給對應的該伺服主機作處理。 The dynamic management system as described in claim 3, wherein the processing module executes steps S1 and S2, and in step S1, when the processing module judges that at least one of the file numbers of the server hosts is not equal to an integer , one of the file quantities is rounded to obtain the integerized file quantity, and in step S2, the processing module substitutes the integerized file quantity into the first formula and the second formula, and then calculates the Rag The Langer multiplier method obtains the number of files to be distributed to each of the rest of the remaining work files of the server host, and the processing module repeatedly executes steps S1 and S2 until each server host Until the number of corresponding files is equal to an integer, the processing module distributes the remaining work files to be processed and equal to the number of files corresponding to each server host to the corresponding server host through the communication module for processing. 一種管理主機,適用於多個伺服主機,並包含一處理模組及電連接該處理模組的一通訊模組,其中,該處理模組經由該通訊模組接收待處理的多個工作檔案(Task),並先將該等工作檔案之其中部分者經由該通訊模組分配給該等 伺服主機作處理,以計算出每一該伺服主機的一運算能力,再根據每一該伺服主機的該運算能力及剩餘待處理的該等工作檔案的數量,將剩餘待處理的該等工作檔案經由該通訊模組分配給該等伺服主機作處理。 A management host, which is suitable for multiple server hosts, and includes a processing module and a communication module electrically connected to the processing module, wherein the processing module receives multiple work files to be processed through the communication module ( Task), and first assign some of these work files to the The server host is processed to calculate a computing power of each server host, and then according to the computing power of each server host and the number of remaining work files to be processed, the remaining work files to be processed Distributed to the server hosts for processing through the communication module. 如請求項5所述的管理主機,其中,該管理主機的該處理模組將所接收且待處理的該等工作檔案依所接收的時間順序排列,並將待處理的該等工作檔案依序分配給該等伺服主機之其中尚未計算出該運算能力者,以獲知每一該伺服主機處理完成單一個該工作檔案所需要的一處理時間,並決定每一該伺服主機的該運算能力等於對應的該處理時間。 The management host as described in claim 5, wherein, the processing module of the management host arranges the received and to-be-processed work files according to the received time order, and sequentially arranges the to-be-processed work files Allocate to one of the server hosts that has not yet calculated the computing power, so as to know the processing time required for each server host to complete a single job file, and determine that the computing power of each server host is equal to the corresponding of the processing time. 如請求項6所述的管理主機,其中,該管理主機的該處理模組根據一第一公式及一第二公式,並以拉格朗日乘數法(Lagrange multiplier)求出要分配給每一該伺服主機的剩餘待處理的該等工作檔案的一檔案數量,該第一公式是:min(該等伺服主機所分別分配的該等檔案數量分別乘以對應的該等運算能力之和),該第二公式是:該等伺服主機所分別分配的該等檔案數量之和=剩餘待處理的該等工作檔案之數量總和,該處理模組再根據每一該伺服主機所對應的該檔案數量,將剩餘待處理的該等工作檔案經由該通訊模組分配給該等伺服主機作處理。 The management host as described in claim 6, wherein, the processing module of the management host calculates the distribution to each A file quantity of the remaining working files of the server host to be processed, the first formula is: min (the number of files allocated by the server hosts is multiplied by the corresponding computing power sum) , the second formula is: the sum of the number of files allocated by the server hosts = the sum of the remaining work files to be processed, and the processing module is based on the file corresponding to each server host Quantity, assigning the remaining work files to be processed to the server hosts through the communication module. 如請求項7所述的管理主機,其中,該處理模組執行步驟S1及S2,於步驟S1,當該處理模組判斷該等伺服主機的該等檔 案數量之其中至少一者不等於整數時,將其中一個該檔案數量作四捨五入計算而獲得整數化的該檔案數量,於步驟S2,該處理模組將整數化的該檔案數量代入該第一公式及該第二公式後再以拉格朗日乘數法求出要分配給其餘的每一該伺服主機的剩餘待處理的該等工作檔案的該檔案數量,該處理模組重覆執行步驟S1及S2,直到每一該伺服主機所對應的該檔案數量都等於整數為止,該處理模組才將剩餘待處理且等於每一該伺服主機所對應的該檔案數量的該等工作檔案,經由該通訊模組分配給對應的該伺服主機作處理。 The management host as described in claim 7, wherein, the processing module executes steps S1 and S2, and in step S1, when the processing module judges that the files of the server hosts When at least one of the file numbers is not equal to an integer, one of the file numbers is rounded to obtain the integerized file number, and in step S2, the processing module substitutes the integerized file number into the first formula and the second formula and then use the Lagrangian multiplier method to obtain the number of files to be allocated to the remaining work files to be processed in each of the remaining server hosts, and the processing module repeatedly executes step S1 and S2, until the number of files corresponding to each of the server hosts is equal to an integer, the processing module will process the remaining work files that are equal to the number of files corresponding to each of the server hosts, through the The communication module is assigned to the corresponding server host for processing.
TW110121375A 2021-06-11 2021-06-11 Dynamic management system and management host TWI785644B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW110121375A TWI785644B (en) 2021-06-11 2021-06-11 Dynamic management system and management host

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW110121375A TWI785644B (en) 2021-06-11 2021-06-11 Dynamic management system and management host

Publications (2)

Publication Number Publication Date
TWI785644B true TWI785644B (en) 2022-12-01
TW202248851A TW202248851A (en) 2022-12-16

Family

ID=85793556

Family Applications (1)

Application Number Title Priority Date Filing Date
TW110121375A TWI785644B (en) 2021-06-11 2021-06-11 Dynamic management system and management host

Country Status (1)

Country Link
TW (1) TWI785644B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020059170A1 (en) * 2000-04-17 2002-05-16 Mark Vange Load balancing between multiple web servers
US20150036493A1 (en) * 2013-07-31 2015-02-05 Citrix Systems, Inc. Systems and methods for load balancing non-ip devices
US20150341407A1 (en) * 2014-05-21 2015-11-26 Sony Corporation System, method, and computer program product for media publishing request processing
TWM600871U (en) * 2020-05-15 2020-09-01 昕力資訊股份有限公司 Computer program product and apparatus for arranging and executing jobs
US20210019188A1 (en) * 2019-07-19 2021-01-21 EMC IP Holding Company LLC Method, electronic device and computer program product for load balance
TWM616624U (en) * 2021-06-11 2021-09-01 中國信託商業銀行股份有限公司 Dynamic management system and management host

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020059170A1 (en) * 2000-04-17 2002-05-16 Mark Vange Load balancing between multiple web servers
US20150036493A1 (en) * 2013-07-31 2015-02-05 Citrix Systems, Inc. Systems and methods for load balancing non-ip devices
US20150341407A1 (en) * 2014-05-21 2015-11-26 Sony Corporation System, method, and computer program product for media publishing request processing
US20210019188A1 (en) * 2019-07-19 2021-01-21 EMC IP Holding Company LLC Method, electronic device and computer program product for load balance
TWM600871U (en) * 2020-05-15 2020-09-01 昕力資訊股份有限公司 Computer program product and apparatus for arranging and executing jobs
TWM616624U (en) * 2021-06-11 2021-09-01 中國信託商業銀行股份有限公司 Dynamic management system and management host

Also Published As

Publication number Publication date
TW202248851A (en) 2022-12-16

Similar Documents

Publication Publication Date Title
WO2018149221A1 (en) Device management method and network management system
US7013344B2 (en) Massively computational parallizable optimization management system and method
US8656387B2 (en) Method and system for workload distributing and processing across a network of replicated virtual machines
JP4959845B2 (en) Method, computer program, and data processing system for calculating dispatcher share in workload manager
US10412154B2 (en) Configuration recommendation for a microservice architecture
CN110795241B (en) Job scheduling management method, scheduling center and system
US20020059427A1 (en) Apparatus and method for dynamically allocating computer resources based on service contract with user
US20080229320A1 (en) Method, an apparatus and a system for controlling of parallel execution of services
CN110597639B (en) CPU distribution control method, device, server and storage medium
EP1654649B1 (en) On demand node and server instance allocation and de-allocation
KR20110060849A (en) Method and system for data distribution in high performance computing cluster
CN111813330A (en) System and method for dispatching input-output
Chaturvedi et al. New heuristic for scheduling of independent tasks in computational grid
US20210191751A1 (en) Method and device for allocating resource in virtualized environment
CN110543352B (en) Task allocation method of scheduling system and related device thereof
US20060031242A1 (en) Method, system, and program for distributing application transactions among work servers
WO2021164265A1 (en) Crowdsourcing task allocation method and apparatus, and electronic device and storage medium
WO2019024235A1 (en) Electronic device, server allocation control method and computer readable storage medium
TWM616624U (en) Dynamic management system and management host
KR20040068103A (en) Apparatus and method of allowing multiple partitions of a partitioned computer system to use a single network adapter
TWI785644B (en) Dynamic management system and management host
KR101678181B1 (en) Parallel processing system
WO2024022142A1 (en) Resource use method and apparatus
CN105208119B (en) A kind of cloud data center method for allocating tasks, device and system
CN110837421A (en) Task allocation method and device