z-logo
open-access-imgOpen Access
Analysis of the use of the Redis in the distributed order processing system in the restaurant network
Author(s) -
V. G. Tkachenko,
Svetlana Lukianiuk
Publication year - 2021
Publication title -
technology audit and production reserves
Language(s) - English
Resource type - Journals
eISSN - 2706-5448
pISSN - 2664-9969
DOI - 10.15587/2706-5448.2021.238460
Subject(s) - computer science , order (exchange) , distributed computing , database , real time computing , economics , finance
The object of research is a distributed order processing system for a restaurant chain. The subject of the research is the analysis of the use of Redis for managing event queues in distributed systems.When implementing a distributed order processing system in a restaurant chain with a possible load of up to 20,000 users per day, the Redis system was used. Management of 9 distributed subsystems was organized through Redis. This solution showed an increase in the performance of the system under heavy load (from 50 transactions per second), but the response time of the system in some cases of its operation was longer than without using Redis. When working systems using Redis, it is necessary to take into account the amount of data with which Redis will work, since it does not exceed the amount of RAM, the absence of differentiation into users and groups, and the absence of a query language, which is replaced by a key-value scheme.This research is aimed at analyzing the operation of the system during trial operation under real load. We compared the operation of a configured system with Redis enabled and disabled. The main indicators for the analysis were the system response time and the maximum request execution time. The research was carried out for 2 weeks, the first week using the system settings with disabled Redis, the second – with enabled Redis. We selected 2 days with a similar load on the system to each other. Especially indicative are the results of comparing the durations of the longest queries, which show an almost constant value of the duration for the system in the mode of enabled Redis. The hypothesis of an increase in the system response time at low loads was confirmed, but this value not only leveled off at a load of 500 unique users but also became less at loads of 1000 unique users.

The content you want is available to Zendy users.

Already have an account? Click here to sign in.
Having issues? You can contact us here