Skip to main content

How to read response time when you run a performance test

By
Measuring the performance of an application is mandatory before releasing it. Measuring the performance of a PoC is also mandatory to validate the base concepts and ideas.
Performance of a system can be measured in different ways, from the process time, to numbers of users, processor level, memory level and so on. Before starting a performance test you should know exactly what you want to measure.
When you want to measure the response time of a specific service/endpoint you should be aware how to interpret the results. Statistical information can be saw in different ways. Each of this view can give you a different perspective of the results.

Average
The average of the response time is calculated. This information is important if you want to know what is the average request time. Even if this information can be very useful, this information is misleading. For example from 1000 requests, you can have an average response time of 16 seconds even if you your chances to have a  requests that takes under 10 seconds is around 90% (you can have 10 requests that takes 300 seconds and 900 requests that takes only 10 seconds).

 Distribution
This view will give you the possibility to see what is the distribution of requests based on the response time. For example you will be able to know that from 1000 requests, 300 requests took 1 seconds, 500 took 2 second, 100 took 9 second and so on.
When you are measuring the scalability of a system, distribution of response is more important than average response time. This information will help you to understand how the requests time change based on different configuration. You can have cases when the average time to be the same, but the distribution of the response time to be very different.
Here we could talk also about mean and standard deviation time.

 Min/Max
When the response time needs to be in a specific time interval (usually a max of X) the min/max of request will offer you this data.

This days I had to measure the performance of a database and how scalable it is. When we started to measure the average execution time of each query with 1, 2 and 3 database nodes we observed that the average response time doesn’t improve so much. In contrast the distribution of query response time is changing a lot. From 30% of requests that take less than 3 second with one node, we ended up with more than 50% of requests that take under 2 seconds when having 3 database nodes.

Labels:

Comments

  1. Cd-MaNDecember 25, 2013 at 12:00 PM

    Related post - why mentioning average/mean and standard deviation is a bad sign :-)

    http://www.javaadvent.com/2013/12/how-not-to-measure-latency.html

    ReplyDelete

Post a Comment

Popular posts from this blog

How to audit an Azure Cosmos DB

By
In this post, we will talk about how we can audit an Azure Cosmos DB database. Before jumping into the problem let us define the business requirement: As an Administrator I want to be able to audit all changes that were done to specific collection inside my Azure Cosmos DB. The requirement is simple, but can be a little tricky to implement fully. First of all when you are using Azure Cosmos DB or any other storage solution there are 99% odds that you’ll have more than one system that writes data to it. This means that you have or not have control on the systems that are doing any create/update/delete operations. Solution 1: Diagnostic Logs Cosmos DB allows us activate diagnostics logs and stream the output a storage account for achieving to other systems like Event Hub or Log Analytics. This would allow us to have information related to who, when, what, response code and how the access operation to our Cosmos DB was done. Beside this there is a field that specifies what was th...
Post a Comment
Read more

Can User Technical Profile Influence AI Architecture Decisions?

By
In the last few days, I ran a small experiment on Devin/Windsurf/Cascade, BMAD, and LLM-based architecture decisions. The initial concern was simple: if an AI coding assistant knows something about the user, for example, that the user is a Java developer or a .NET developer, can this information influence the architecture and technology stack that the AI will propose? This question is important because many companies are starting to use AI tools not only for coding but also for product discovery, architecture, documentation, and technical decision-making. If the AI is influenced by the developer's personal profile, the result may not be fully neutral. It may look like a business or architecture decision, but in reality, it may be partially shaped by the context of the person using the tool. The test was done using BMAD with Devin/Windsurf/Cascade. The scenario was a unified field service operations platform. The task was to let BMAD refine the product, define the architecture and s...
Post a Comment
Read more

ADO.NET provider with invariant name 'System.Data.SqlClient' could not be loaded

By
Today blog post will be started with the following error when running DB tests on the CI machine: threw exception: System.InvalidOperationException: The Entity Framework provider type 'System.Data.Entity.SqlServer.SqlProviderServices, EntityFramework.SqlServer' registered in the application config file for the ADO.NET provider with invariant name 'System.Data.SqlClient' could not be loaded. Make sure that the assembly-qualified name is used and that the assembly is available to the running application. See http://go.microsoft.com/fwlink/?LinkId=260882 for more information. at System.Data.Entity.Infrastructure.DependencyResolution.ProviderServicesFactory.GetInstance(String providerTypeName, String providerInvariantName) This error happened only on the Continuous Integration machine. On the devs machines, everything has fine. The classic problem – on my machine it’s working. The CI has the following configuration: TeamCity .NET 4.51 EF 6.0.2 VS2013 It see...
8 comments
Read more