: How to test the speed of a page that is guarded behind login page? For public page, I can use pingdom.com to calculate the response time/uptime of a page over a certain period of time.

@Carla537

You can use google chrome to do this

Open up the developers tools (Press Command+Option+I (Mac) or Control+Shift+I (Windows, Linux))

Go to the Network tab.

Refresh the page.

You will see a list of every resource that your page loads.
Once finished you can see the total time it took to load every resource at the bottom

I have checked this after login at my Gmail. Like this screenshot:


There is a "disable cache" option that can be useful to see loading time of the first time a page loads, compared to a repeated visitor.

Vote Up Vote Down

@Lengel546

2017 update

The Lighthouse tool developed by Google can be run as a Chrome extension on logged-in pages, and even against Chrome on a real mobile device (which you should use instead of emulators whenever possible). Lighthouse provides audits for performance, accessibility, progressive web apps, and more.

Here's a screenshot of Lighthouse auditing a Google Keep note page:



In addition to using the extension against logged-in pages, you can do it on the CLI, as well.

Vote Up Vote Down

@Yeniel560

If what you want to know is response time of the site for a first request, or sever response for starting the handshake, ping is most than enough since that is independent from the content retrieved, the handshake or the headers. It only depends on ICMP requests being allowed.

If you need to calculate loading time, that is different, the best way to do that, and considering that you mention that the page/site is personalized, is to create a page with the same amount and kind of content, create specific no cache rules on your .htaccess and the use whichever tool you want to test response and loading time.

Of course that page has to do the same work that the normal application does, just with non sensitive data, for that, you should create some fake data on the DB id the is any associated, then the page has to be generated with the same server side technology if you are using any for the normal pages and it has to fetch and load the same images.

Of course that is if you want accurate data. If you don't want to do that work, or you don't need data as accurate as the one you will get with that procedure, then you can use tools like ab command from Apache or any other simulation tool.

Vote Up Vote Down

@Gloria169

You can also install a browser extension such as Yslow or PageSpeed Insights and run the test directly from your browser.

Vote Up Vote Down

@Mendez628

One possible approach is to create a backdoor that works only for whitelisted IP addresses. Pingdom's IP range, for example, is available on their site, so you could temporarily allow requests from that IP range through without authentication.

It's possible to spoof the requesting IP address, but since the requester won't receive any data back from your server (it'll be sent to the IP they're spoofing), this approach might be secure enough for your purposes. Of course, it's best not to leave a backdoor open when you don't need it.

Vote Up Vote Down

@Welton855

Google Analytics now has Site Speed information. You can run the analytics script on logged-in pages and view the results there.

Vote Up Vote Down

@Carla537

Create a backdoor (just remember to remove it later on!)

For example: Access through page inbox.php?loginForTest=1 will automatically login you as a user with ID 25, or whichever you'd like to use use for test, and load this content without making any further authentication checks.

Assuming you write OOP you'd be able to change just a few lines of code in authentication class to load the content from a certain user perspective if certain GET data is set.

As for HTTP Authentication - you got most of the answers in comments already. Use the @ syntax in URL to login as a user.

If you still won't be able to get through the login form - try to use firebug for more basic tests - just remember to run in several times to get a proper average and don't compare data from firebug with data from pingdom or other similar services as they might have different method of obtaining timings.

Vote Up Vote Down