[go: up one dir, main page]

JP2008003736A - How to warn of insufficient computer resources - Google Patents

How to warn of insufficient computer resources Download PDF

Info

Publication number
JP2008003736A
JP2008003736A JP2006170864A JP2006170864A JP2008003736A JP 2008003736 A JP2008003736 A JP 2008003736A JP 2006170864 A JP2006170864 A JP 2006170864A JP 2006170864 A JP2006170864 A JP 2006170864A JP 2008003736 A JP2008003736 A JP 2008003736A
Authority
JP
Japan
Prior art keywords
job execution
computer
prediction
job
time
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP2006170864A
Other languages
Japanese (ja)
Inventor
Tsuneo Kikuchi
恒男 菊池
Yuko Tanaka
雄孝 田中
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hitachi Ltd
Original Assignee
Hitachi Ltd
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 Hitachi Ltd filed Critical Hitachi Ltd
Priority to JP2006170864A priority Critical patent/JP2008003736A/en
Publication of JP2008003736A publication Critical patent/JP2008003736A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

【課題】利用者にジョブ実行用計算機を提供する計算機センタにおいて、利用者が計算機資源量を上回るジョブ実行を行いたい場合、計算機資源量が制限になり、全てが実行できず、業務の遅延が発生する。本発明の課題は、計算機資源の不足が発生しないようにすることにより、業務遅延を回避することにある。
【解決手段】利用者が申請するジョブ実行予定と、ジョブ実行実績から、ジョブ実行予測を行い、今後計算機資源追加に掛かる日数分の間にジョブ実行予測が計算機資源を上回ることが予測された場合には警告する。
【選択図】図2
[PROBLEMS] In a computer center that provides a computer for job execution to a user, if the user wants to execute a job that exceeds the amount of computer resources, the amount of computer resources is limited, all of them cannot be executed, and work delays occur. appear. An object of the present invention is to avoid a business delay by preventing a shortage of computer resources.
When a job execution prediction is made based on a job execution schedule applied by a user and a job execution result, and it is predicted that the job execution prediction will exceed the computer resources for the number of days required for adding computer resources in the future. To warn.
[Selection] Figure 2

Description

本発明は、ジョブ実行用計算機を利用者に提供する計算機センタにおいて、利用者から申請されたジョブ実行予定と、ジョブ実行実績から、今後のジョブ実行を予測し、利用者に割り当てたジョブ実行用計算機のジョブ実行可能時間合計をジョブ実行予測が上回る場合に警告する方法に関する。   The present invention predicts future job execution from a job execution schedule requested by a user and a job execution result in a computer center that provides the user with a job execution computer. The present invention relates to a method for warning when a job execution prediction exceeds the total job execution time of a computer.

複数の計算機が存在する計算機センタにおいて、利用者は実行予定を事前提出し、計算機センタ管理者は実行予定に基づき計算機資源を準備する。利用者が見積もりを誤り実行予定以上にジョブ実行しようとした場合、計算機資源が制限となり、ジョブ実行は利用者が実行しようとしたジョブ全てが実行出来ず、利用者の業務が遅延する。   In a computer center having a plurality of computers, a user submits an execution schedule in advance, and a computer center administrator prepares computer resources based on the execution schedule. If a user tries to execute a job more than the estimated execution error, the computer resources will be limited, and the job execution will not be able to execute all the jobs that the user has attempted to execute, and the user's work will be delayed.

また、利用者が実行しようとしたジョブ全てが実行出来ないということが判明した後に、計算機資源の追加対策には数日必要となり、対策に掛かる日数は、利用者の業務が遅延する。   Further, after it is found that all the jobs that the user tried to execute cannot be executed, additional measures for computer resources are required for several days, and the number of days required for the measures delays the user's work.

特開2005−250818号公報JP-A-2005-250818

本発明は、計算機資源の不足が発生しないようにすることにより、業務遅延を回避する手段を提供する。   The present invention provides means for avoiding business delays by preventing a shortage of computer resources.

利用者が事前に提出した実行予定と、実行実績から、計算機資源の追加導入に掛かる日数分の実行予測を行い、その予測値が計算機センタの計算機資源を超過する場合には警告を発する。この警告に基づき、計算機資源の不足が実際に発生する前に計算機資源の追加導入作業を開始し、計算機資源の不足が発生することを防ぐ。   Based on the execution schedule submitted by the user and the execution results, execution prediction is performed for the number of days required for additional introduction of computer resources, and a warning is issued when the predicted value exceeds the computer resources of the computer center. Based on this warning, the computer resource additional introduction work is started before the computer resource shortage actually occurs to prevent the computer resource shortage from occurring.

本発明によれば、計算機資源の不足が発生しないようにすることにより、業務遅延を回避することができる。   According to the present invention, it is possible to avoid a business delay by preventing a shortage of computer resources.

以下本発明を実施例に基づいて説明する。   Hereinafter, the present invention will be described based on examples.

図1は本発明を実施する計算機センタの機器構成図の一例である。計算機を使用する利用者はジョブ実行用計算機101でジョブ実行する。管理用計算機103は、各々のジョブ実行用計算機101で実行されたジョブのジョブ実行時間を、ネットワーク102を介して集め、計算機資源の今後の実行予測を行い、計算機資源の不足が予測された場合は警告を発する。   FIG. 1 is an example of a device configuration diagram of a computer center for implementing the present invention. A user who uses the computer executes a job on the job execution computer 101. When the management computer 103 collects job execution times of jobs executed by the respective job execution computers 101 via the network 102 and performs future prediction of computer resources, and a shortage of computer resources is predicted. Issues a warning.

図2は計算機資源を予測するシステムのブロック図である。計算機を使用するユーザは、事前にジョブ実行予定(日毎のジョブ時間)を申請し、本データはジョブ実行予定データ201に登録しておく。計算機センタ管理者は事前に下記(1)(2)を計算機センタ仕様データ202へ登録しておく。(1)計算機資源追加を決めてから作業完了するまでの日数(以降、資源追加日数とする)。(2)利用者に割り当てたジョブ実行用計算機のジョブ実行可能時間合計(以降、計算機資源量とする)。ジョブ実行時間収集機能203は、ジョブ実行用計算機で実行された1日のジョブ実行時間を収集し、ジョブ実行実績データ204に出力する。使用計算機資源予測機能205は、ジョブ実行予定データ201と、ジョブ実行実績データ204から、今後の使用計算機資源量を予測し、計算機センタ仕様データ202に登録されている計算機資源量を上回る場合、警告メッセージ表示206によりアナウンスされる。   FIG. 2 is a block diagram of a system for predicting computer resources. A user who uses the computer applies for a job execution schedule (job time for each day) in advance, and this data is registered in the job execution schedule data 201. The computer center manager registers the following (1) and (2) in the computer center specification data 202 in advance. (1) The number of days from the decision to add computer resources to the completion of the work (hereinafter referred to as resource addition days). (2) The total job executable time of the computer for job execution assigned to the user (hereinafter referred to as computer resource amount). The job execution time collection function 203 collects the daily job execution time executed by the job execution computer and outputs it to the job execution result data 204. The used computer resource prediction function 205 predicts the future computer resource amount from the job execution schedule data 201 and the job execution result data 204, and warns when the computer resource amount registered in the computer center specification data 202 exceeds the computer resource amount. Announced by message display 206.

図3は使用計算機資源予測機能205の第一例を示す図である。実線の折れ線301は利用者が事前提出したジョブ実行予定を示し、一線鎖線の直線304は計算機資源量を示す。実線の棒グラフは利用者が実際に使用したジョブ実行時間を示し、点線の棒グラフ303は予測されたジョブ実行予測時間である。予測方法を説明する。現時点でのジョブ実行実績302が110時間であり、同時点での実行予定が100時間である場合、この予定と実績の比を計算する(予定・実績比)。本例の場合110÷100=1.1となる。次の日以降も利用者がジョブ実行予定に対して予定・実績比で使用すると予測し、次の日以降のジョブ実行予定に対して予定・実績比を積算し、ジョブ実行予測303を決定する。この予測を資源追加日数分計算し、ジョブ実行予測303が計算機資源量304を上回る日がある場合に警告メッセージを出力する。   FIG. 3 is a diagram showing a first example of the used computer resource prediction function 205. A solid line 301 indicates a job execution schedule submitted in advance by the user, and a chain line 304 indicates a computer resource amount. A solid bar graph indicates the job execution time actually used by the user, and a dotted bar graph 303 indicates the predicted job execution time. A prediction method will be described. When the current job execution record 302 is 110 hours and the execution schedule at the same time is 100 hours, the ratio of this schedule to the record is calculated (schedule / result ratio). In this example, 110 ÷ 100 = 1.1. From the next day onward, it is predicted that the user will use the planned / actual ratio with respect to the job execution schedule, and the job execution prediction 303 is determined by adding the planned / actual ratio to the job execution schedule on and after the next day. . This prediction is calculated for the number of days of resource addition, and a warning message is output when there is a date when the job execution prediction 303 exceeds the computer resource amount 304.

図4は使用計算機資源予測機能205の第一例の処理フローである。ジョブ実行予定に対するジョブ実行実績の比を計算し(ステップ401)(計算結果をAとする)、資源追加日数までのジョブ実行予定に、ステップ401で計算された値(A)を乗算する(ステップ402)(計算結果をBとする)。ステップ402で計算された値(B)が計算機資源量を上回るか比較し(ステップ403)、上回る場合は警告メッセージを表示する(ステップ404)。   FIG. 4 is a processing flow of a first example of the used computer resource prediction function 205. The ratio of the job execution result to the job execution schedule is calculated (step 401) (the calculation result is A), and the job execution schedule up to the number of resource addition days is multiplied by the value (A) calculated in step 401 (step). 402) (the calculation result is B). Whether the value (B) calculated in step 402 exceeds the computer resource amount is compared (step 403), and if it exceeds, a warning message is displayed (step 404).

図5は実施例1で述べた使用計算機資源予測機能205の第二例を示す図である。実線の折れ線501は利用者が事前提出したジョブ実行予定を示し、一線鎖線の直線504は計算機資源量を示す。実線の棒グラフは利用者が実際に使用したジョブ実行時間を示し、点線の棒グラフ503は予測されたジョブ実行予測時間である。予測方法を説明する。現時点でのジョブ実行実績502が110時間であり、同時点での実行予定が100時間である場合、この予定と実績の差を計算する(予定・実績差)。本例の場合110−100=10時間となる。次の日以降もジョブ実行予定に対して予定・実績差で利用者が使用すると予測し、次の日以降のジョブ実行予定に対する予定・実績差を加算し、ジョブ実行予測503を決定する。この予測を資源追加日数分計算し、ジョブ実行予測503が計算機資源量504を上回る日がある場合に警告メッセージを出力する。   FIG. 5 is a diagram illustrating a second example of the used computer resource prediction function 205 described in the first embodiment. A solid line 501 indicates a job execution schedule submitted in advance by the user, and a chain line 504 indicates a computer resource amount. A solid bar graph indicates the job execution time actually used by the user, and a dotted bar graph 503 indicates the predicted job execution time. A prediction method will be described. If the current job execution result 502 is 110 hours and the execution schedule at the same time is 100 hours, the difference between this schedule and the result is calculated (schedule / result difference). In this example, 110-100 = 10 hours. From the next day onward, it is predicted that the user will use the planned / actual difference with respect to the job execution schedule, and the job execution prediction 503 is determined by adding the planned / actual difference to the job execution schedule on and after the next day. This prediction is calculated for the number of days of resource addition, and a warning message is output when there is a day when the job execution prediction 503 exceeds the computer resource amount 504.

図6は使用計算機資源予測機能205の第二例の処理フローである。ジョブ実行予定に対するジョブ実行実績の差を計算し(ステップ601)(計算結果をCとする)。資源追加日数までのジョブ実行予定に、ステップ601で計算された値(C)を加算する(ステップ602)(計算結果をDとする)。ステップ602で計算された値(D)が計算機資源量を上回るか比較し(ステップ603)、上回る場合は警告メッセージを表示する(ステップ604)。   FIG. 6 is a processing flow of a second example of the used computer resource prediction function 205. The difference between the job execution results and the job execution schedule is calculated (step 601) (the calculation result is C). The value (C) calculated in step 601 is added to the job execution schedule up to the resource addition days (step 602) (the calculation result is D). Whether the value (D) calculated in step 602 exceeds the computer resource amount is compared (step 603), and if it exceeds, a warning message is displayed (step 604).

図7は実施例1に述べた使用計算機資源予測機能205の第三例である。利用者のジョブ実行は(1)、(2)に示す特徴を持つ場合がある。(1)業務における総ジョブ実行時間が決まっている。(2)業務の終了日は遅延出来ない。この業務においては、開始後業務日程までのジョブ実行時間がジョブ実行予定より少ない場合、以降の日程の中で予定より多いジョブ実行時間を要することとなる。このことを利用したジョブ実行予測を図7により説明する。実線の折れ線701は利用者が事前提出したジョブ実行予定を示し、一線鎖線の直線709は計算機資源量を示す。実線の棒グラフは利用者が実際に使用したジョブ実行時間を示し、点線の曲線708は予測されたジョブ実行予測時間である。予測方法を説明する。業務開始時から現時点までのジョブ実行予定701に対するジョブ実行実績702のジョブ実行時間差の累計を計算する。これをジョブ実行予定・実績差累計とし、図中の面積で表すと703となる。次に現時点から業務完了までのジョブ実行予定701のジョブ時間の累計を計算する。図中の面積で表すと704となる。ここで次に示す(1)〜(4)を満足する楕円の軌道をジョブ実行時間予測とする。(1)中心を通る横軸がジョブ時間0の軸705である。(2)現時点のジョブ実行実績時間の棒グラフ頂点706を通る。(3)業務終了時のジョブ時間0の点707を通る。(4)現時点から業務完了までの楕円708と、ジョブ時間0の軸705で囲まれる部分の面積が、ジョブ実行予定・実績差累計703とジョブ実行予定時間累計704を加算したものと等しい。このジョブ実行時間予測となる楕円709が、資源追加日数分において、計算機資源量709を上回る場合に警告メッセージを出力する。本例で資源追加日数を3日とすると、現時点706の一日後、二日後のジョブ実行予測は計算機資源量709を下回るが、三日後のジョブ実行予測710が計算機資源量709を上回る。このとき警告メッセージが出力され、計算機資源追加の対策を行う。   FIG. 7 shows a third example of the used computer resource prediction function 205 described in the first embodiment. A user's job execution may have the characteristics shown in (1) and (2). (1) The total job execution time in business is determined. (2) The work end date cannot be delayed. In this business, when the job execution time from the start to the business schedule is less than the job execution schedule, the job execution time longer than the schedule is required in the subsequent schedules. Job execution prediction using this fact will be described with reference to FIG. A solid line 701 indicates a job execution schedule submitted in advance by the user, and a chain line 709 indicates a computer resource amount. A solid bar graph indicates the job execution time actually used by the user, and a dotted curve 708 indicates the predicted job execution time. A prediction method will be described. The cumulative total of job execution time differences of the job execution results 702 with respect to the job execution schedule 701 from the start of business to the current time is calculated. This is the job execution schedule / actual difference cumulative total, and is expressed as 703 in the area in the figure. Next, the total job time of the job execution schedule 701 from the current time to the completion of the job is calculated. In terms of the area in the figure, 704 is obtained. Here, an elliptical trajectory satisfying the following (1) to (4) is set as the job execution time prediction. (1) A horizontal axis passing through the center is an axis 705 of 0 job time. (2) A bar graph vertex 706 of the current job execution result time is passed. (3) Passes a point 707 of job time 0 at the end of the job. (4) The area surrounded by the ellipse 708 from the current time to the completion of the work and the axis 705 of job time 0 is equal to the sum of the job execution schedule / result difference cumulative 703 and the job execution scheduled time cumulative 704. A warning message is output when the ellipse 709 serving as the job execution time prediction exceeds the computer resource amount 709 in the number of days of resource addition. In this example, assuming that the number of days of resource addition is 3, the job execution prediction after one day and two days after the current time 706 is less than the computer resource amount 709, but the job execution prediction 710 after three days exceeds the computer resource amount 709. At this time, a warning message is output and countermeasures for adding computer resources are taken.

図8は使用計算機資源予測機能205の第三例の処理フローである。業務開始時から現時点までのジョブ実行予定に対するジョブ実行実績の差を累計し(ステップ801)(計算結果をEとする)、現時点から業務完了までのジョブ実行予定のジョブ時間累計を計算する(ステップ802)(計算結果をFとする)。図7で説明した楕円軌道を計算する(ステップ803)。このとき楕円の軌道で示される現時点から業務終了までのジョブ実行予測時間の累計はEとFを加算したものとなる。資源追加日数までの楕円で示されるジョブ実行予測時間を算出し(ステップ804)(計算結果をGとする)。ステップ804で計算された値(G)が計算機資源量を上回るか比較し(ステップ805)、上回る場合は警告メッセージを表示する(ステップ806)。   FIG. 8 is a processing flow of the third example of the used computer resource prediction function 205. The difference between job execution results with respect to the job execution schedule from the start of business to the current time is accumulated (step 801) (the calculation result is E), and the cumulative job time of the job execution schedule from the current time to the completion of business is calculated (step) 802) (the calculation result is F). The elliptical trajectory described in FIG. 7 is calculated (step 803). At this time, the cumulative total job execution time from the present time to the end of the work indicated by an elliptical trajectory is obtained by adding E and F. The estimated job execution time indicated by an ellipse up to the resource addition days is calculated (step 804) (the calculation result is G). Whether the value (G) calculated in step 804 exceeds the computer resource amount is compared (step 805), and if it exceeds, a warning message is displayed (step 806).

ジョブ実行業務において過去の統計から、当日に実行するジョブ実行時間は、前日のジョブ実行時間と関連付けが可能である。前日に対する当日のジョブ実行時間の変異率を、ジョブ実行時間変移率とする。このジョブ実行時間変移率を利用して、ジョブ実行予測処理の必要・不要が判断することができる。   Based on past statistics in job execution work, the job execution time to be executed on the current day can be related to the job execution time of the previous day. The change rate of the job execution time on the current day relative to the previous day is defined as the job execution time change rate. Using this job execution time change rate, the necessity / unnecessity of job execution prediction processing can be determined.

図9は前日と当日のジョブ実行時間変移率を度数分布で表した調査結果である。過去統計から変移率の最大値が110%であると算出出来た場合、これからジョブ実行予測処理の必要・不要が判断できる。例として資源追加日数を3日、計算機資源量を500時間とすると、資源追加日数が経過しても計算機資源量を上回らないジョブ実行実績は500[時間]÷110[%]の3乗=375.65[時間]と計算できる。本値を余裕実行限度時間とする。ジョブ実行実績が375時間だった場合、今後の実行時間が最大の変移率110%で増加して行くとした場合の資源追加日数3日後のジョブ実行時間は、375[時間]×110[%]の3乗=499[時間]と計算でき計算機資源量500時間を下回る。つまりジョブ実行実績が余裕実行限度時間を下回る場合は、資源追加日数内で計算機資源量を超える可能性が無いと考え、ジョブ実行予測処理は不要とできる。   FIG. 9 shows the results of a survey showing the job execution time change rate of the previous day and the current day as a frequency distribution. When it is possible to calculate from the past statistics that the maximum value of the transition rate is 110%, it is possible to determine whether or not the job execution prediction process is necessary. As an example, assuming that the number of days for adding resources is 3 days and the amount of computer resources is 500 hours, the job execution result that does not exceed the amount of computer resources even if the number of days of resource addition elapses is the cube of 500 [hours] / 110 [%] = 375 .65 [hours]. This value is the marginal execution limit time. When the job execution result is 375 hours, the job execution time after 3 days of resource addition when the future execution time increases at the maximum change rate of 110% is 375 [hour] × 110 [%] Can be calculated as cube of 3 = 499 [hours], and the amount of computer resources is less than 500 hours. That is, when the job execution result is less than the marginal execution limit time, it is considered that there is no possibility of exceeding the computer resource amount within the resource addition days, and the job execution prediction process can be made unnecessary.

図10はジョブ実行実績データを元に、ジョブ実行予測必要・不要を判断する機能を持たせたジョブ実行予測を行うシステムの構成図である。計算機を使用するユーザは、事前にジョブ実行予定(日毎のジョブ時間)を申請し、本データはジョブ実行予定データ1001に登録しておく。計算機センタ管理者は事前に下記(1)(2)を計算機センタ仕様データ1002へ登録しておく。(1)計算機資源追加を決めてから作業完了するまでの日数(資源追加日数)。(2)利用者に割り当てたジョブ実行用計算機が持っているジョブ実行可能時間合計(計算機資源量)。ジョブ実行時間収集機能1003は、ジョブ実行用計算機で実行された1日のジョブ実行時間を収集し、ジョブ実行実績データ1004に出力する。余裕実行限度時間計算機能1005は、ジョブ実行実績データ1004からジョブ実行時間変移率の最大値を算出し、本値と計算機センタ仕様データ1002の計算機資源量から余裕実行限度時間を算出し、余裕実行限度時間データ1006へ登録する。予測要・不要判断付き使用計算機資源予測機能1007は、ジョブ余裕実行限界時間データ1006とジョブ実行実績データ1004を比較することによりジョブ実行予測必要・不要を判断し、必要な場合はジョブ使用時間予測を行い、計算機センタ仕様データ1002に登録されている計算機資源量を上回る場合、警告メッセージ表示1008でアナウンスする。   FIG. 10 is a configuration diagram of a system that performs job execution prediction based on job execution result data and has a function of determining whether or not job execution prediction is necessary. A user who uses the computer applies for a job execution schedule (job time for each day) in advance, and this data is registered in the job execution schedule data 1001. The computer center manager registers the following (1) and (2) in the computer center specification data 1002 in advance. (1) The number of days from the decision to add computer resources to the completion of the work (the number of days to add resources). (2) Total job executable time (computer resource amount) possessed by the job execution computer assigned to the user. The job execution time collection function 1003 collects the daily job execution time executed by the job execution computer and outputs the collected job execution time data 1004. The marginal execution limit time calculation function 1005 calculates the maximum value of the job execution time transition rate from the job execution result data 1004, calculates the marginal execution limit time from this value and the computer resource amount of the computer center specification data 1002, and executes the marginal execution. It is registered in the limit time data 1006. The computer resource prediction function 1007 with prediction necessity / unnecessary judgment judges whether or not job execution prediction is necessary / unnecessary by comparing the job margin execution limit time data 1006 with the job execution result data 1004, and if necessary, predicts the job usage time. When the amount of computer resources registered in the computer center specification data 1002 is exceeded, a warning message display 1008 is announced.

図11は予測要・不要判断付き使用計算機資源予測機能1003の処理フローである。ジョブ実行実績と余裕実行限界時間を比較し(ステップ1101)、ジョブ実行実績が上回る場合は使用計算機資源予測機能1102によりジョブ実行予測を行い、計算機資源量を超える場合は警告メッセージ表示を行う。本例では使用計算機資源予測機能1102の内容は、実施例1のジョブ実行予定と実績の比からジョブ実行予測を行い警告する処理となっている。使用計算機資源予測機能1102の内容は、実施例2、実施例3で示した処理方法のいずれも利用できる。   FIG. 11 is a processing flow of the used computer resource prediction function 1003 with prediction necessity / unnecessity determination. The job execution result and the marginal execution limit time are compared (step 1101). If the job execution result exceeds, the job execution prediction is performed by the used computer resource prediction function 1102, and if the computer resource amount is exceeded, a warning message is displayed. In this example, the content of the used computer resource prediction function 1102 is a process for performing a job execution prediction from the ratio of the job execution schedule and the actual result of the first embodiment and giving a warning. For the contents of the used computer resource prediction function 1102, any of the processing methods shown in the second and third embodiments can be used.

計算機センタの機器構成図である。It is an apparatus block diagram of a computer center. 本発明の一実施例である計算機資源不足を警告するシステムの構成図である。1 is a configuration diagram of a system that warns of a shortage of computer resources according to an embodiment of the present invention. FIG. ジョブ実行予定と実績の比からジョブ実行予測を行い警告する処理の説明図である。It is explanatory drawing of the process which performs job execution prediction from the ratio of a job execution schedule and a performance, and warns. ジョブ実行予定と実績の比からジョブ実行予測を行い警告する処理フローである。This is a processing flow in which a job execution prediction is made based on a ratio between a job execution schedule and a result and a warning is given. ジョブ実行予定と実績の差からジョブ実行予測を行い警告する処理の説明図である。It is explanatory drawing of the process which makes job execution prediction from the difference of a job execution schedule and a performance, and warns. ジョブ実行予定と実績の差からジョブ実行予測を行い警告する処理フローである。This is a processing flow in which a job execution prediction is made from the difference between the job execution schedule and the actual result and a warning is given. ジョブ実行時間の累計からジョブ実行予測を行い警告する処理の説明図である。It is explanatory drawing of the process which warns by performing job execution prediction from the accumulation of job execution time. ジョブ実行時間の累計からジョブ実行予測を行い警告する処理フローである。This is a processing flow in which job execution prediction is made from the accumulated job execution time and a warning is issued. ジョブ実行実績の日毎の変異率の測定結果である。It is a measurement result of the daily variation rate of job execution results. 本発明の一実施例であるジョブ実行予測の要・不要を判断する手段を備えた計算機資源不足を警告するシステムの構成図である。1 is a configuration diagram of a system that warns of a shortage of computer resources provided with a means for determining necessity / unnecessity of job execution prediction according to an embodiment of the present invention. FIG. ジョブ実行予測の要・不要を判断する手段を備えたジョブ実行予測を行い警告する処理フローである。It is a processing flow for performing job prediction and providing a warning with means for determining whether or not job execution prediction is necessary.

符号の説明Explanation of symbols

101…ジョブ実行用計算機、102…ネットワーク、103…管理用計算機、201…ジョブ実行予定データ、202…計算機センタ仕様データ、203…ジョブ実行時間収集機能、204…ジョブ実行実績データ、205…使用計算機資源予測機能、206…警告メッセージ表示機能、301…ジョブ実行予定、302…ジョブ実行実績、303…ジョブ実行予測、304…計算機資源量、501…ジョブ実行予定、502…ジョブ実行実績、503…ジョブ実行予測、504…計算機資源量、701…ジョブ実行予定、702…ジョブ実行実績、703…現時点までのジョブ実行予定とジョブ実行実績の差、704…現時点以降のジョブ実行予定、705…グラフ横軸、706…現時点のジョブ実行実績時間、707…業務終了日、708…ジョブ実行予測、709…計算機資源量、710…現時点から3日後のジョブ実行予測、1001…ジョブ実行予定データ、1002…計算機センタ仕様データ。
DESCRIPTION OF SYMBOLS 101 ... Job execution computer, 102 ... Network, 103 ... Management computer, 201 ... Job execution scheduled data, 202 ... Computer center specification data, 203 ... Job execution time collection function, 204 ... Job execution result data, 205 ... Used computer Resource prediction function, 206 ... Warning message display function, 301 ... Job execution schedule, 302 ... Job execution record, 303 ... Job execution prediction, 304 ... Computer resource amount, 501 ... Job execution schedule, 502 ... Job execution record, 503 ... Job Execution prediction, 504 ... computer resource amount, 701 ... job execution schedule, 702 ... job execution record, 703 ... difference between job execution schedule and job execution record up to now, 704 ... job execution schedule after this point, 705 ... horizontal axis of graph 706 ... Current job execution result time 707 ... Business end date, 7 8 ... job execution prediction, 709 ... computer resource amount, the job execution prediction after 3 days from 710 ... Currently, 1001 ... job execution schedule data, 1002 ... computer center specification data.

Claims (5)

計算機利用者のジョブ実行予定とジョブ実行実績の比からジョブ実行予測を行い、計算機資源量を上回る場合に警告する方法。   A method that predicts job execution based on the ratio between the computer user's job execution schedule and job execution results, and warns when the amount of computer resources is exceeded. 計算機利用者のジョブ実行予定とジョブ実行実績の差からジョブ実行予測を行い、計算機資源量を上回る場合に警告する方法。   A method that predicts job execution from the difference between the computer user's job execution schedule and job execution results, and warns when the amount of computer resources is exceeded. 現時点までの計算機利用者のジョブ実行予定とジョブ実行実績の差の累計と、現時点以降のジョブ実行予定累計の加算からジョブ実行予測を行い、計算機資源量を上回る場合に警告する方法。   A method for predicting job execution based on the sum of the difference between the job execution schedule and job execution results of the computer user up to the present time and adding the cumulative job execution schedule after the current time, and warns when the amount of computer resources is exceeded. ジョブ実行実績の変移率からジョブ実行予測の要・不要を判断する手段を備えた請求項1、もしくは請求項2、もしくは請求項3の警告する方法。   The warning method according to claim 1, 2, or 3, further comprising means for determining whether or not job execution prediction is necessary or not from a change rate of job execution results. 請求項1から請求項4の方法を組み合わせて提供する警告の方法。
5. A warning method provided by combining the methods of claim 1 to claim 4.
JP2006170864A 2006-06-21 2006-06-21 How to warn of insufficient computer resources Pending JP2008003736A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2006170864A JP2008003736A (en) 2006-06-21 2006-06-21 How to warn of insufficient computer resources

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2006170864A JP2008003736A (en) 2006-06-21 2006-06-21 How to warn of insufficient computer resources

Publications (1)

Publication Number Publication Date
JP2008003736A true JP2008003736A (en) 2008-01-10

Family

ID=39008074

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2006170864A Pending JP2008003736A (en) 2006-06-21 2006-06-21 How to warn of insufficient computer resources

Country Status (1)

Country Link
JP (1) JP2008003736A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009217373A (en) * 2008-03-07 2009-09-24 Ns Solutions Corp Information processor, information processing method and program
JP2010066828A (en) * 2008-09-08 2010-03-25 Ns Solutions Corp Information processor, information processing method and program
EP4040298A1 (en) 2021-02-09 2022-08-10 Fujitsu Limited Method of calculating predicted exhaustion date and program of calculating predicted exhaustion date
JPWO2023162000A1 (en) * 2022-02-22 2023-08-31

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009217373A (en) * 2008-03-07 2009-09-24 Ns Solutions Corp Information processor, information processing method and program
JP2010066828A (en) * 2008-09-08 2010-03-25 Ns Solutions Corp Information processor, information processing method and program
EP4040298A1 (en) 2021-02-09 2022-08-10 Fujitsu Limited Method of calculating predicted exhaustion date and program of calculating predicted exhaustion date
JPWO2023162000A1 (en) * 2022-02-22 2023-08-31
WO2023162000A1 (en) * 2022-02-22 2023-08-31 日本電信電話株式会社 Resource determination device, method, and program
JP7768337B2 (en) 2022-02-22 2025-11-12 Ntt株式会社 Resource determination device, method, and program

Similar Documents

Publication Publication Date Title
US8560667B2 (en) Analysis method and apparatus
Moore et al. Transforming reactive auto-scaling into proactive auto-scaling
EP1282038A2 (en) Distributed processing system and distributed job processing method
US10783002B1 (en) Cost determination of a service call
US20110138397A1 (en) Processing time estimation method and apparatus
JP4977729B2 (en) Job schedule change support system, method and program
JP3936924B2 (en) Job scheduling method and system
CN119376890A (en) Task resource scheduling method, device, equipment and storage medium
US12135996B2 (en) Computing resource scheduling method, scheduler, internet of things system, and computer readable medium
US20090254411A1 (en) System and method for automated decision support for service transition management
Wierman et al. Scheduling despite inexact job-size information
CN114840392B (en) Task scheduling anomaly monitoring method, device, medium and program product
CN114064204A (en) A method for dynamic expansion based on business prediction in a microservice environment
Choi et al. An enhanced data-locality-aware task scheduling algorithm for hadoop applications
CN108491255B (en) Self-service MapReduce data optimal distribution method and system
US20090240476A1 (en) Method and device for forecasting computational needs of an application
CN103442087A (en) Web service system access volume control device and method based on response time trend analysis
Maglaras et al. Queueing systems with leadtime constraints: A fluid-model approach for admission and sequencing control
JP2005099973A (en) Operation management system
JP2008003736A (en) How to warn of insufficient computer resources
JP4705484B2 (en) Performance monitoring device, performance monitoring method and program
Iglesias et al. Increasing task consolidation efficiency by using more accurate resource estimations
JP2018041296A (en) Computer system and job execution plan change method
Makhijani et al. Quest: Queue simulation for content moderation at scale
Birkenheuer et al. The gain of overbooking