MA37726A1 - Technical for the safety of calculation grids - Google Patents

Technical for the safety of calculation grids

Info

Publication number
MA37726A1
MA37726A1 MA37726A MA37726A MA37726A1 MA 37726 A1 MA37726 A1 MA 37726A1 MA 37726 A MA37726 A MA 37726A MA 37726 A MA37726 A MA 37726A MA 37726 A1 MA37726 A1 MA 37726A1
Authority
MA
Morocco
Prior art keywords
credibility
reputation
broker
resource
tracking
Prior art date
Application number
MA37726A
Other languages
French (fr)
Inventor
Essaaidi Mohamed
Bendahmane Ahmed
Original Assignee
Université Mohammed V De Rabat
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 Université Mohammed V De Rabat filed Critical Université Mohammed V De Rabat
Priority to MA37726A priority Critical patent/MA37726A1/en
Publication of MA37726A1 publication Critical patent/MA37726A1/en

Links

Landscapes

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

Abstract

Dans cet article, nous proposons une nouvelle approche à base de vote réputation, basé sur la technique de contrôles ponctuels. Cette approche améliore du vote basé la crédibilité pour atteindre un taux d'erreur faible, avec un surcout raisonnable. Dans le vote basé sur la crédibilité, un contrôle ponctuel est utilisé pour vérifier périodiquement les ressources informatiques, par l'envoi d’un travail de repérage dont le résultat correct est connu, afin d'estimer la crédibilité de chaque ressource basée sur le résultat retourné. Cette technique conduit à la dissipation des ressources dans le schéma de vote simple, car il nécessite des calculs supplémentaires pour produire le résultat des travaux de repérage. L'idée de base de notre approche proposée consiste à vérifier les ressources informatiques sans attribuer l’emploi de repérage et d'envisager la suite de la décision de vote que celui des contrôles inopinés pour estimer la crédibilité sans calculs supplémentaires. Cette crédibilité est considérée comme réputation qui est utilisée dans la décision, selon l’algorithme de la figure 1. Dans cet algorithme, le service de courtier de grille commence le calcul par l'envoi d’un certain nombre de a ressources de calcul disponibles à la ligne (2) pour la transformation parallélisée. Apres le calcul, les résultats générés seront retournés au courtier (lignes 19 et 20). Tant que toutes les taches sont terminées avec des résultats acceptés (lignes 4-7), la programmation et les processus de réception pour chaque tache sont répétés. Supposons que chaque tâche est reproduite n fois et attribuée à plusieurs ressources informatiques ci, de sorte qu'un courtier peut percevoir des résultats différents vj m, ou i = 1, 2, . . ., n et j = 1, 2, . . ., m. Chaque ressource informatique a sa réputation la valeur ri, qui représentent le comportement global. Cette réputation est recueillie par un service broker- grille, qui contient la liste de la réputation de toutes les ressources de calcul. La réputation est une valeur dans la plage entre 0 et 1. Selon notre vision, le service de courtier de grille construit la réputation de chaque ressource informatique grâce à sa crédibilité. La crédibilité représente la probabilité d’un objet particulier du système à fonctionner correctement. En général, la crédibilité cr (ci, ki) de la ressource informatique ci est calculée en passant des contrôles sur place le temps de ki. Puisque nous considérons chaque tache validée avec succès par le courtier en utilisant la rbv comme une tache de repérage.In this article, we propose a new approach based on vote reputation, based on the technique of spot checks. This approach improves vote-based credibility to achieve a low error rate, with a reasonable surcharge. In the vote based on credibility, a spot check is used to periodically check the computer resources, by sending a tracking job whose correct result is known, in order to estimate the credibility of each resource based on the result return. This technique leads to the dissipation of resources in the simple voting scheme, because it requires additional calculations to produce the result of the tracking works. The basic idea of our proposed approach is to check computing resources without assigning the use of tracking and consider following the voting decision as that of unannounced checks to estimate credibility without further calculations. This credibility is considered reputation that is used in the decision, according to the algorithm of Figure 1. In this algorithm, the grid broker service starts the calculation by sending a number of available computing resources in line (2) for the parallelized transformation. After the calculation, the generated results will be returned to the broker (lines 19 and 20). As long as all the tasks are completed with accepted results (lines 4-7), the programming and reception processes for each task are repeated. Suppose that each task is reproduced n times and allocated to several computing resources ci, so that a broker can perceive different results vj m, where i = 1, 2,. . ., n and j = 1, 2,. . ., m. Each IT resource has its reputation the value ri, which represents the overall behavior. This reputation is collected by a broker-grid service, which contains the list of the reputation of all computing resources. The reputation is a value in the range between 0 and 1. According to our vision, the grid broker service builds the reputation of each IT resource through its credibility. Credibility represents the probability of a particular object of the system working properly. In general, the credibility cr (ci, ki) of the computing resource ci is calculated by passing on-the-spot checks the time of ki. Since we consider each task successfully validated by the broker using the rbv as a tracking spot.

MA37726A 2014-12-31 2014-12-31 Technical for the safety of calculation grids MA37726A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
MA37726A MA37726A1 (en) 2014-12-31 2014-12-31 Technical for the safety of calculation grids

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
MA37726A MA37726A1 (en) 2014-12-31 2014-12-31 Technical for the safety of calculation grids

Publications (1)

Publication Number Publication Date
MA37726A1 true MA37726A1 (en) 2016-07-29

Family

ID=57197719

Family Applications (1)

Application Number Title Priority Date Filing Date
MA37726A MA37726A1 (en) 2014-12-31 2014-12-31 Technical for the safety of calculation grids

Country Status (1)

Country Link
MA (1) MA37726A1 (en)

Similar Documents

Publication Publication Date Title
US12217196B2 (en) Facilitating device fingerprinting through assignment of fuzzy device identifiers
AR116476A1 (en) METHOD AND SYSTEM TO EXECUTE AUTOMATIC LEARNING ALGORITHMS
TW201944338A (en) Data processing method, device, equipment and server for identifying insurance fraud
EA202091986A8 (en) METHOD FOR SEARCHING OR COMPARING POINTS USING ROUTES IN THE TRANSPORTATION SYSTEM
Lin et al. Covariance structure regularization via entropy loss function
JP2021047414A (en) Device and computer readable storage medium for voice fingerprint collation
CN111383005A (en) Digital currency flow direction tracking method and device
CN103116582A (en) Information retrieval method and relevant system and device
CN112395320B (en) Building information merging method, device, equipment and computer readable storage medium
Yang et al. A probabilistic model for truth discovery with object correlations
US10870432B1 (en) Systems and methods for assessing vehicle operation in a direction of sunlight
CN107688563B (en) Method and device for identifying synonyms
Orihara Stock market listing and corporate policy: Evidence from reforms to Japanese corporate law
MA37726A1 (en) Technical for the safety of calculation grids
Toumi et al. AI for climate change: unveiling pathways to sustainable development through greenhouse gas emission predictions
Lekkas et al. Finite mixture models in neighbourhoods-to-health research: A systematic review
CN104636318A (en) Distributed or increment calculation method of big data variance and standard deviation
CN111858927B (en) Data testing method and device, electronic equipment and storage medium
JPWO2019069507A1 (en) Feature generator, feature generator and feature generator
CN114154477A (en) Text data processing method and device, electronic equipment and readable storage medium
Lee et al. Web Scraping Crawling-based Automatic Data Augmentation for Deep Neural Networks-based Vehicle Classifications
Corso Toward predictive crime analysis via social media, big data, and gis spatial correlation
US20150262196A1 (en) Electronic Financial/Economic Modeling Environment
US20250200123A1 (en) ENHANCED SEARCH REPORT BASED ON LEVERAGING SEARCH ENGINE APIs, LARGE LANGUAGE MODELS, AND WEB CRAWLERS
Hasanaj et al. Cooperative edge deepfake detection