JPH09134328A - Method for evaluating stability of computer system - Google Patents

Method for evaluating stability of computer system

Info

Publication number
JPH09134328A
JPH09134328A JP28903495A JP28903495A JPH09134328A JP H09134328 A JPH09134328 A JP H09134328A JP 28903495 A JP28903495 A JP 28903495A JP 28903495 A JP28903495 A JP 28903495A JP H09134328 A JPH09134328 A JP H09134328A
Authority
JP
Japan
Prior art keywords
stability
computer network
flow
equation
computer
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
JP28903495A
Other languages
Japanese (ja)
Inventor
Masami Hiramatsu
真美 平松
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.)
Meidensha Corp
Meidensha Electric Manufacturing Co Ltd
Original Assignee
Meidensha Corp
Meidensha Electric Manufacturing Co 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 Meidensha Corp, Meidensha Electric Manufacturing Co Ltd filed Critical Meidensha Corp
Priority to JP28903495A priority Critical patent/JPH09134328A/en
Publication of JPH09134328A publication Critical patent/JPH09134328A/en
Pending legal-status Critical Current

Links

Abstract

PROBLEM TO BE SOLVED: To evaluate stability by permitting the flow of information between respective computer elements to correspond to the flow of substances when respective producers in a specified ecology model are preyed and obtaining the stability of a system from a specified expression. SOLUTION: At the time of constructing a hierarchical computer network system, the flow of information between the respective compute elements constituting the system is permitted to correspond to the flow of the substances when the respective producers in the Lotka-Vorlterra ecology model are preyed. Moreover, the stability of the system is obtained from the expression compared with the equation of the ecology model. Thus, at the time of constructing the computer network system, its stability is evaluated. Besides, the optimum arrangement computer network is surely and easily constructed. The computer network system suitable to needs is constructed in accordance with the extension and reduction of the system.

Description

【発明の詳細な説明】Detailed Description of the Invention

【0001】[0001]

【発明の属する技術分野】本発明は、階層型コンピュー
タネットワークシステムを構築するためのコンピュータ
システムの安定性評価方法に関する。
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer system stability evaluation method for constructing a hierarchical computer network system.

【0002】[0002]

【従来の技術】コンピュータネットワークは、コンピュ
ータや端末を相互に接続し、演算能力やソフトウェア、
データベースなどの情報資源を共有し、相互に利用する
ことを可能とする。
2. Description of the Related Art A computer network connects computers and terminals to each other to provide computing power, software,
Information resources such as databases can be shared and used mutually.

【0003】従来のコンピュータネットワークシステム
の構築方法としては、図3に示すように個々のワークス
テーションを並列にLANで結んだネットワークや、図
4に示すように各クライアントへのサービスをサーバを
介して行うクライアント・サーバ方式がある。
As a conventional method for constructing a computer network system, as shown in FIG. 3, a network in which individual workstations are connected in parallel with each other by a LAN, or, as shown in FIG. 4, services for clients are provided via a server. There is a client / server method.

【0004】[0004]

【発明が解決しようとする課題】従来のコンピュータネ
ットワークシステムの構築方法は、システムの構築目的
に基づいて必要な資源を分散してその有効利用ができる
ように構築されるため、システムの拡張性が低くなる場
合が多い。また、システムを拡張可能とした場合にもそ
の安定性が得られるかどうかの評価がなされておらず、
拡張性が保証されるものでもないし、安定したシステム
の構築に失敗することがある。
Since the conventional method for constructing a computer network system is constructed so that necessary resources can be distributed and utilized effectively based on the purpose of constructing the system, system scalability is improved. It often becomes low. Also, even if the system is expandable, it has not been evaluated whether the stability is obtained,
Scalability is not guaranteed and stable system construction may fail.

【0005】本発明の目的は、安定性を評価したシステ
ムの構築及び拡張ができるコンピュータシステムの安定
性評価方法を提供することにある。
An object of the present invention is to provide a stability evaluation method for a computer system, which enables construction and expansion of a stability evaluated system.

【0006】[0006]

【課題を解決するための手段】本発明は、階層型コンピ
ュータネットワークシステムを構築するにおいて、前記
システムを構成する各コンピュータ要素間の情報の流れ
をLotka−Volterra生態モデルの各生産者
が捕食される際の物質の流れに対応させ、前記生態モデ
ルの方程式に対比させた次の式、
According to the present invention, in constructing a hierarchical computer network system, each producer of the Lotka-Volterra ecological model eats the flow of information between the computer elements constituting the system. The following equation, which corresponds to the flow of matter at the time and is compared with the equation of the ecological model,

【0007】[0007]

【数2】 (Equation 2)

【0008】からシステムの安定度を求めることを特徴
とする。
The stability of the system is obtained from the above.

【0009】[0009]

【発明の実施の形態】図1は、階層型コンピュータネッ
トワークシステムと生態モデルを対比させて示す。同図
の右側に示す階層型システムを構築するのに、左側に示
す生態システムのモデルの1つであるLotka−Vo
lterraモデル(捕食関係、階層型)をコンピュー
タネットワークの各コンピュータの階層関係にあてはめ
る。
FIG. 1 shows a hierarchical computer network system and an ecological model in contrast. To construct the hierarchical system shown on the right side of the figure, one of the models of the ecosystem shown on the left side, Lotka-Vo.
The lterra model (predatory relation, hierarchical type) is applied to the hierarchical relation of each computer in the computer network.

【0010】Lotka−Volterraモデルは、
生態システムの各生産者が捕食される際の物質の流れを
示し、この物質の流れを垂直、水平分散の階層型コンピ
ュータネットワーク内での情報の流れに対応させること
でシステムの安定性の評価を行う。
The Lotka-Volterra model is
The stability of the system is evaluated by showing the flow of substances when each producer of the ecosystem is predated and correlating the flow of substances with the flow of information in a vertically and horizontally distributed hierarchical computer network. To do.

【0011】Lotka−Volterraモデルの方
程式は、各生産者を要素とし、その固体数や自己増殖、
要素間の影響などをあてはめると次式で表現される。
The equation of the Lotka-Volterra model has each producer as an element, and the number of individuals and self-reproduction,
When the influence between elements is applied, it is expressed by the following equation.

【0012】[0012]

【数3】 (Equation 3)

【0013】この方程式を階層型コンピュータネットワ
ークシステムにあてはめると、各コンピュータを要素と
し、そのプロセス数や生成率・消滅率等で生態プロセス
に対応させると次の式を得ることができる。
When this equation is applied to a hierarchical computer network system, the following equation can be obtained by using each computer as an element and making it correspond to an ecological process by the number of processes, the generation rate, the extinction rate, etc.

【0014】[0014]

【数4】 (Equation 4)

【0015】本実施形態では、上記の方程式からシステ
ムの安定性の評価を行う。この評価は、最初にシステム
のコンピュータの台数、階層数を決め、図2に示す手順
でなされる。
In this embodiment, the stability of the system is evaluated from the above equation. This evaluation is performed by the procedure shown in FIG. 2 after first determining the number of computers and the number of layers of the system.

【0016】(S1)システム構成の中から1つのパタ
ーンを取り出す。
(S1) One pattern is extracted from the system configuration.

【0017】(S2)コンピュータネットワークシステ
ムの各要素について上記の(2)式に示す各パラメータ
を決める。
(S2) Each parameter shown in the above equation (2) is determined for each element of the computer network system.

【0018】(S3)パラメータが設定された(2)式
におけるシステムの平衡点を求める。この平衡点は、決
められたパターンについてのプロセス数Ii*について
の変化率が零になる次の式から求められる。
(S3) Find the equilibrium point of the system in the equation (2) in which the parameters are set. This equilibrium point is obtained from the following equation in which the rate of change for the number of processes I i * for a given pattern becomes zero.

【0019】[0019]

【数5】dt/dIi*=0 …(3) つまり、Dt / dI i * = 0 (3) That is,

【0020】[0020]

【数6】 (Equation 6)

【0021】を解いてIi*を求める。Solving for I i *.

【0022】(S4)求めたIi*がi=1〜nの1つ
でも負であるか否かをチェックする。I1*に1つで負
があるときは、このネットワークは局所不安定が存在
し、安定な状態では存在できないと見做す。
(S4) It is checked whether or not the calculated I i * is negative even if one of i = 1 to n. When I 1 * is 1 and is negative, the network is considered to have local instability and cannot exist in a stable state.

【0023】(S5)上記の平衡点演算でIi*の全て
が負の数でないとき、全ての要素が平衡点近傍で局所安
定かどうかを調べる。これには、上記の(2)式を次の
(5)式のように線形化する。
(S5) In the above equilibrium point calculation, when all of I i * are not negative numbers, it is checked whether all the elements are locally stable in the vicinity of the equilibrium point. For this purpose, the equation (2) is linearized as the following equation (5).

【0024】[0024]

【数7】dI/dt=FI …(5) F={fiji*|i,j=1,..,n} (S6)上記の線形化による行列Fの固有値の実部が全
て負か否かをチェックする。このチェックで1つでも負
にならないときは局所不安定とし、すべて負になるとき
は局所安定とする。このときの局所安定度TRを以下の
ように定義する。
## EQU00007 ## dI / dt = FI (5) F = {f ij I i * | i, j = 1 ,. . , N} (S6) It is checked whether all real parts of the eigenvalues of the matrix F obtained by the above linearization are negative. If even one of these checks is not negative, it is considered locally unstable, and if all are negative, then it is locally stable. The local stability TR at this time is defined as follows.

【0025】TR;局所安定であるときの行列Fの固有
値の最大値の逆数。
TR: The reciprocal of the maximum value of the eigenvalues of the matrix F when it is locally stable.

【0026】この安定度TRは、収束に最も影響を与え
る時定数であり、局所安定になる速さと見做すことがで
きる。したがって、TRの値が小さいほど安定性が高い
と見做す。
The stability TR is a time constant that most affects the convergence, and can be regarded as the speed at which the stability becomes local. Therefore, the smaller the value of TR, the higher the stability.

【0027】(S7)上記までの手順を各パターンにつ
いて演算する。
(S7) The procedure up to the above is calculated for each pattern.

【0028】(S8)各パターンについての演算が終了
したとき、最初に設定したコンピュータの台数で可能な
限りのパターンについての安定度TRの中で最も小さい
パターンを最適なネットワークの配置として決定する。
(S8) When the calculation for each pattern is completed, the smallest pattern among the stability TRs for the patterns possible with the initially set number of computers is determined as the optimum network arrangement.

【0029】[0029]

【発明の効果】以上のとおり、本発明によれば、システ
ムを構成する各コンピュータ要素間の情報の流れをLo
tka−Volterra生態モデルの各生産者が捕食
される際の物質の流れに対応させ、生態モデルの方程式
に対比させた式からシステムの安定度を求めるようにし
たため、コンピュータネットワークシステムの構築に際
してその安定性を評価できる。
As described above, according to the present invention, the flow of information between the computer elements constituting the system is Lo.
Since the producers of the tka-Volterra ecological model corresponded to the flow of substances during predation, the stability of the system was determined from the equations contrasted with the equations of the ecological model. You can evaluate the sex.

【0030】また、最適な配置のコンピュータネットワ
ークを構築するのを確実、容易にする。
It also ensures and facilitates the construction of an optimally arranged computer network.

【0031】また、システムの拡張、縮小に対応してニ
ーズに合ったコンピュータネットワークシステムを構築
できる。
Further, it is possible to construct a computer network system which meets the needs in correspondence with the expansion and reduction of the system.

【図面の簡単な説明】[Brief description of the drawings]

【図1】本発明の実施形態を示す生態モデルとコンピュ
ータネットワークの対比を示す図。
FIG. 1 is a diagram showing a comparison between an ecological model showing an embodiment of the present invention and a computer network.

【図2】実施形態における安定性評価手順図。FIG. 2 is a stability evaluation procedure diagram in the embodiment.

【図3】従来の並列LAN方式のコンピュータネットワ
ーク。
FIG. 3 is a conventional parallel LAN computer network.

【図4】従来のクライアント・サーバ方式のコンピュー
タネットワーク。
FIG. 4 is a conventional client-server computer network.

Claims (1)

【特許請求の範囲】[Claims] 【請求項1】 階層型コンピュータネットワークシステ
ムを構築するにおいて、 前記システムを構成する各コンピュータ要素間の情報の
流れをLotka−Volterra生態モデルの各生
産者が捕食される際の物質の流れに対応させ、前記生態
モデルの方程式に対比させた次の式、 【数1】 からシステムの安定度を求めることを特徴とするコンピ
ュータシステムの安定性評価方法。
1. In constructing a hierarchical computer network system, the flow of information between computer elements constituting the system is made to correspond to the flow of substances when each producer of the Lotka-Volterra ecological model is eaten. , The following equation in contrast to the equation of the ecological model, A method for evaluating the stability of a computer system, characterized in that the stability of the system is obtained from the method.
JP28903495A 1995-11-08 1995-11-08 Method for evaluating stability of computer system Pending JPH09134328A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP28903495A JPH09134328A (en) 1995-11-08 1995-11-08 Method for evaluating stability of computer system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP28903495A JPH09134328A (en) 1995-11-08 1995-11-08 Method for evaluating stability of computer system

Publications (1)

Publication Number Publication Date
JPH09134328A true JPH09134328A (en) 1997-05-20

Family

ID=17737978

Family Applications (1)

Application Number Title Priority Date Filing Date
JP28903495A Pending JPH09134328A (en) 1995-11-08 1995-11-08 Method for evaluating stability of computer system

Country Status (1)

Country Link
JP (1) JPH09134328A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110032146A (en) * 2019-04-24 2019-07-19 西安交通大学 A kind of complicated processing process stability appraisal procedure based on the multi-machine collaborative factor

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110032146A (en) * 2019-04-24 2019-07-19 西安交通大学 A kind of complicated processing process stability appraisal procedure based on the multi-machine collaborative factor
CN110032146B (en) * 2019-04-24 2020-10-27 西安交通大学 Complex machining process stability evaluation method based on multi-machine synergistic factors

Similar Documents

Publication Publication Date Title
Chen et al. Quantile regression under memory constraint
Li et al. MSGD: A novel matrix factorization approach for large-scale collaborative filtering recommender systems on GPUs
Biloš et al. Neural flows: Efficient alternative to neural ODEs
Khosravi et al. On tractable computation of expected predictions
CN111243045A (en) Image generation method based on Gaussian mixture model prior variation self-encoder
Liu et al. Privacy-preserving incentive mechanism design for federated cloud-edge learning
TWI700599B (en) Method and device for embedding relationship network diagram, computer readable storage medium and computing equipment
CN111342991B (en) Information propagation method based on cross-social network
Huang et al. Global stability for epidemic models on multiplex networks
CN105786626B (en) The Storm platform thread distribution method divided based on K
JPH09134328A (en) Method for evaluating stability of computer system
Buccafurri et al. Experimenting with certified reputation in a competitive multi-agent scenario
Lv et al. Global convergence of Oja's PCA learning algorithm with a non-zero-approaching adaptive learning rate
JP2005242580A (en) Parameter estimation method, data prediction method, parameter estimation device, data prediction device, and computer program
CN117892805B (en) Personalized federal learning method based on supernetwork and hierarchy collaborative graph aggregation
US11861490B1 (en) Decoupled machine learning training
Mollasalehi et al. Modelling and analysis of relation between load balancing and scalability in distributed computing systems
Cobb et al. Approximating probability density functions with mixtures of truncated exponentials
WO2023233931A1 (en) Design assistance device and design assistance method
JPH06325136A (en) Numerical analysis device
Peltomäki et al. Characterizing spatiotemporal patterns in three-state lattice models
Wenisch et al. Harnessing high-performance computers for computational steering
CN106331696A (en) Frame loss detection method and device
JP2004102831A (en) Parameter estimation method, data prediction method, parameter estimation device, data prediction device, computer program, and recording medium
Kolaczyk et al. Mathematical Models for Network Graphs