Methodology for calculating the load, the number of users of an information system - a website or service
When developing / creating a web site, a mobile application, a WEB service — in other words, an Information System (IS), the question arises of requiring hardware resources — the number of servers (virtual machines).
The above methodology describes the calculation of the number of users and "equipment" for the territory of the Russian Federation.
It would be nice to have a more detailed diagram of the functional structure of the system, in its absence - although a simple description of the modules and their interaction.
Users - the target audience is a percentage of all residents of Russia. Variable adjustable when calculating / modeling.
The hourly load is distributed as follows:
The parameters of daily distribution - can also be adjusted during modeling.
')
Basic indicators - population size, according to Rosstat "According to Rosstat" Population size of the Russian Federation by municipalities as of January 1, 2012, thousand people, from here onwards to Rosstat. Normalize, remove doubles and occurrences, add to the table the time zone in accordance with the PP (government decree) No. 725 dated August 31, 2011 or in WiKi - Time Zones of Russia .
The resulting table (hereinafter, the drawings from the Excel file - the original file is available on GoogleDoc).
and further - let's not clutter the article.
The result of the distribution data is a Pivot table.
By the way, pay attention, an unexpected (at least for me) result - 101 million people live ... Moscow time.
Next, worksheet (fragment)
Management of calculation of approximate numerical indicators at 2 levels
For the system described in the example, the final result looks like this:
It is clear that the system is created quite simple as the simplest user profile - performing simple operations.
It is also necessary to keep in mind and be able to use analytical reports for forecasts, for example - BRANCH REPORT. Federal Agency for Press and Mass Communications Internet in Russia State, trends and development prospects .
or
Mobile Internet Russia .
It was interesting to simulate the calculated model on a more complex user profile and system - to improve the methodology.
The above methodology describes the calculation of the number of users and "equipment" for the territory of the Russian Federation.
Initial data
- Website "Business card";
- WEB-service for mobile applications, working under the protocol http / https, interacting with the Database;
- SQL Database (NOSQL);
- A WEB client is a mobile application implementing functionality for web users, which also interacts with the Database.
It would be nice to have a more detailed diagram of the functional structure of the system, in its absence - although a simple description of the modules and their interaction.
Users - the target audience is a percentage of all residents of Russia. Variable adjustable when calculating / modeling.
The hourly load is distributed as follows:
- Peak from 8 o'clock local time rise to 70% within an hour;
- Reducing the load up to 18 hours to 50%;
- Reducing the load up to 22 hours to 10%;
- Reducing the load up to 00 hours to a level of 5% before the morning peak.
The parameters of daily distribution - can also be adjusted during modeling.
')
Proceed to the calculation
Basic indicators - population size, according to Rosstat "According to Rosstat" Population size of the Russian Federation by municipalities as of January 1, 2012, thousand people, from here onwards to Rosstat. Normalize, remove doubles and occurrences, add to the table the time zone in accordance with the PP (government decree) No. 725 dated August 31, 2011 or in WiKi - Time Zones of Russia .
The resulting table (hereinafter, the drawings from the Excel file - the original file is available on GoogleDoc).
and further - let's not clutter the article.
The result of the distribution data is a Pivot table.
By the way, pay attention, an unexpected (at least for me) result - 101 million people live ... Moscow time.
Next, worksheet (fragment)
Management of calculation of approximate numerical indicators at 2 levels
- Directly set the number of users of the System (with the ratio of our users 100%)
- The size of the target group (for example, 22 million households in Russia) and the percentage of their coverage in the column “our users”
For the system described in the example, the final result looks like this:
It is clear that the system is created quite simple as the simplest user profile - performing simple operations.
It is also necessary to keep in mind and be able to use analytical reports for forecasts, for example - BRANCH REPORT. Federal Agency for Press and Mass Communications Internet in Russia State, trends and development prospects .
or
Mobile Internet Russia .
It was interesting to simulate the calculated model on a more complex user profile and system - to improve the methodology.
Source: https://habr.com/ru/post/187386/
All Articles