- Author: [[Jack Rothrock]] - Full Title: The 7 Application Performance Metrics You Need to Measure and Why | Scout APM Blog - Tags:: - URL: https://scoutapm.com/blog/application-performance-metrics-you-need-why - ### Highlights first synced by [[Readwise]] [[2020-12-19]] - 3. Queue Time When we talk about Queue Time, we have to recognize that this is a measurable side effect of Little’s Law. - > Little’s law is a queueing theory/law that can be used to understand potential bottlenecks in said queues. There are a few queues that most developers need to be cognizant about when it comes to an application’s performance: requests flowing from load balancers to web servers and background job worker queues. - > Little’s Law is an algebraic expression that describes, in relation to web applications, the amount of transactions in a system is equal to the arrival rate (throughput) multiplied by the average time in the system (average response time). - Using this we can rework the equation to figure out what happens when either throughput goes up or response time goes up. With load balancers, seeing an increase in queue time is an indication that the servers are running out of resources. Since queue time is just a measurable effect of Little’s Law, we can use Little’s Law to cure queue time. One way is to reduce average response time via increasing performance or doing less work per request, the other is to have a higher throughput tolerance via adding more servers. - # Zettels - [[Little's Law for calculating transactions in a system]] - [[Queue Time]] -