当前位置:网站首页>What is little's law

What is little's law

2021-05-04 14:02:59 Jiedao jdon

Anyone involved in agile and Kanban can't avoid Little The laws of (Little's Law),Little The law is an equation : 

L = λ W

The variable here means :

L = The average number of tasks or items in a queuing system over a period of time

λ= The average number of new tasks or projects entering the system within a specified time interval ( Arrival rate

= The average time spent by a task or project throughout the system

The Chinese meaning of this equation is :

“  The number of items in the queue  ”=“  New project arrival rate  ”x“  The average time spent on a task or project  ”

First you need to know what a queuing system is ?Little Our law only applies to “ Queuing system ”, It's a queue system that has to have entry and exit . It could be a workflow system , It could be mission systems , It could be a production line .

In software development , Tasks or projects are also often referred to as user stories , Change request , Error repair, etc .

“L” Represents the number of items in the queuing system you are checking . This is also called “WIP”, Such as “ In progress ” Project , It can be almost any integer .

“λ” Represents the arrival rate and departure rate of items entering and leaving the queuing system . This is sometimes called “ throughput ” or “ Access and / Or the amount of items leaving the system ”, And sometimes expressed as λ or “A”.

Arrival rates are a bit confusing at first , But the key to remember is that it's usually just a small part . This is because you want to measure the entry of items into / Out of the system rate , Not the number of items or the time between new arrivals . therefore ,“λ” Always expressed as a fraction :

λ = ( A project ) / ( Time unit )

for example , If every new project 20 Minutes into your queue , Then your arrival rate is not 20, It is 1/20.

Last ,“W” It's the average time an item spends in a queuing system . This is also called “ Lead time ”, It can also be any unit of time . The time unit of this element needs to be the same as “λ” In the same time unit - If you measure the arrival rate in days , that “W” It will also be measured in days , And so on .

The final meaning is :

The number of items in the system   =( Entry and exit rates )x( The average time a project spends in the system )

L = λ W  By algebraic transformation to :   

W  =  L / λ 

As I said before L yes WIP,  λ Can be throughput , So a variation of this formula is as follows :

LT  =  WIP  /   throughput

LT = Average delivery time

WIP =  The average number of tasks in the queue ( Work in progress )

throughput   = The average number of tasks that leave the system in a defined time interval

problem : Under the assumption that throughput remains unchanged , If we keep increasing WIP, That is, the number of items in the system , Whether the delivery time keeps increasing ?

Queuing The theory points out that , As the utilization rate increases, more than 80%, The speed of the Internet has dropped dramatically ( nonlinear ). But according to Little Law ( Given a stable system ), If we add WIP( Improve utilization ), Then the time will increase linearly . Why? Little's Law There's no indication that delivery times are rising exponentially from a certain point ?

Reference resources :Little’s Law: Isn’t It a Linear Relationship?

 

版权声明
本文为[Jiedao jdon]所创,转载请带上原文链接,感谢
https://chowdera.com/2021/05/20210504135746363y.html

随机推荐