Page response baseline testing

🗄️ Archive
1

Hi everyone,

The project I will be testing soon is introducing Cloud Application Security (CAS). What i need to test is not web page performance as such, more any increase in time for users to access/use a SaaS before and after the CAS implementation.

The devices I will test on are:

  • PC (win7/10)
  • Managed devices (iPhone and Andriod)
  • Unmanaged Devices (iPhone and Andriod)
    [I can go into further detail on the devices matrix if needed]

Does any have any recommendations for tools I could use?

What is your opinion on using Chrome Dev Tools lighthouse? Can this be used for web page response times, or is it really just for the performance of the web page itself?

Thanks :slight_smile:

2

Hi,

We are measuring loading times for our product and we are using the developer tools and a web proxy and monitoring tool called Charles https://www.charlesproxy.com/. If you can or cannot use developer tools depends a lot on you actual application. In our case we load the application in several stages and then have continued communication between the client and the server which makes it a little harder to say “this was the loading time”. But I tried for this page and it has an initial load of 3.27 s.
Charles is mainly used to measure mobiles since you just setup a proxy in Charles and configure your client to connect to that proxy instead. Then you can see all the traffic and timings from the clients.

I will also monitor this thread to see if anyone else have other suggestions as we still would like the process easier on our end.

Good Luck!

3

For those interested, we ended up setting up Selenium and using Node.js performance.mark([name])

const {PerformanceObserver, performance} = require(‘perf_hooks’);

As we were in a situation where we could easily change between the existing and new web proxy, this allowed us to measure the time a page takes to load at very close to the same time of day/user load.

Simplistic javescript to show performance marks:

await driver.get(‘https://google.com’);
performance.mark(‘InitialMark’);

//Load page
await driver.get(‘https://www.cisco.com/’);

performance.mark(‘FinalMark’);
performance.measure(‘Total Duration’,‘InitialMark’,‘FinalMark’);

I expanded other test cases to go through other workflows, such as sharing a file in SharePoint

1 Like