Assessment of the distance learning server\'s operation strategies and service capacity in advance
Download 399.25 Kb. Pdf ko'rish
|
e3sconf ebwff2023 06016
- Bu sahifa navigatsiya:
- Server with 4 service devices Requests sent at the same time Requests served Rejected request
Fig. 1. Changes in the level of server utilization in relation to the intensity of requests
0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0% 5 10 14 16 18 20 22 24 Ser ver ut iliz at io n r at e (ρ ) λ - speed of requests entering the system (in second) E3S Web of Conferences 420, 06016 (2023) https://doi.org/10.1051/e3sconf/202342006016 EBWFF 2023 5 Due to the rise in requests, the service time was also examined, and the results are shown in Fig. 2. Fig. 2. The M/M/m model's variation of service time upon intensity However, it should also be emphasized that this model assumes a situation where there are no queues on the system server, and it will not be acceptable for systems with queues. This indicates that the working principle of the systems whose service time is established based on (4) is based on the transparent loss technique, and when all the service devices are busy during the system service, the request to the system will not be served. They also affected the server's service time when this view was served, as service devices in the system generated a load on the system during the period up to the loss of requests that arrived when it was busy. Determining the presence of an empty device Server with 4 service devices Requests sent at the same time Requests served Rejected request Rejecting the request due to the fact that all service devices are busy Fig. 3. The process of providing services in the transparent loss method There are also methods that work with servers based on the principle of waiting, and the M/G/m/k model was used to determine the service time of systems based on this principle in relation to the number of requests. In this service method, the average response time to requests was calculated by the manner of queuing on the server [26]. In this method, the distance learning server was studied in connection with one request service time (SRT), the rate of arrival of requests at different values. Formula (5) was used to determine the average service time of the M/G/m/k model. Also, the estimation of the average service time based on this model depends on various factors, the relevant values of which were determined based on the following formulas: (6), (7), and (8). ܴܶ ൌ ଵ ሺ כ ఓ ି ఒሻ ା ௐ (5) Here, W q is the average waiting time in the queue, which is determined using the following formula (6). 0 2 4 6 8 10 12 14 16 5 10 14 16 18 20 22 24 Download 399.25 Kb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling