0

We have our development page giving correct score on page speed insights.

https://developers.google.com/speed/pagespeed/insights/?hl=en&url=https%3A%2F%2Fdev.techtitute.com

But our production page always returns scan error:

Lighthouse returned error: FAILED_DOCUMENT_REQUEST. Lighthouse was unable to reliably load the page you requested. Make sure you are testing the correct URL and that the server is properly responding to all requests. (Details: net::ERR_CONNECTION_FAILED)

https://developers.google.com/speed/pagespeed/insights/?hl=en&url=https%3A%2F%2Fwww.techtitute.com

The strange thing is that web.dev (which also uses Lighthouse under the hoods) is giving correct score for both pages, also gtmetrix, pingdom and Chrome dev tools Lighthouse tab.

Both pages have the exact same code, both are running on windows Azure App Service. The only differences are:

On production we are injecting Google Tag Manager scripts with analytics, fb pixel, linkedin and others. Also we have an azure traffic manager to distribute traffic based on geolocalization of the user (Europe or Latam).

What would you think could be the problem of production page giving that error?

UPDATE 09/07/2020

We have created a parallel environment of traffic manager geographically distributed pointing to 2 app services with the same code as production, but with another domain: dev3.techtitute.com. Now page speed is giving the same error on this subdomain, as the production one. So we think it's related to azure traffic manager configuration.

  • I would suggest checking what happens if the user is outside of the geolocation areas you stated. I couldn't see anything wrong that was immediately obvious (other than a 502 error, but both the dev site and the production site have that problem so it is almost certainly not that). My guess is Google is being blocked or running into a redirect loop etc. etc. The geolocation is the first place I would look as their servers are probably US based for Page Speed Insights (but we don't actually know this). See if you have anything in your error logs. One of those tough ones to track down. – GrahamTheDev Jul 08 '20 at 23:01
  • Thank you for your comment, we have tested the web through VPN servers all over the world and there are no visible problems. Also we have tested a specific page without loading Google Tag Manager and all the scripts, the error continues the same (everything ok in dev, and fails in production). What 502 error did you see in both dev and production sites? – David Lorenzo López Jul 09 '20 at 09:29
  • 1
    I looked at it from germany. Works fine. But the movie in the background ist loading for 1,3 minutes. Could be that you are hitting a timeout. – snitch182 Jul 09 '20 at 12:12
  • Thanks for your comments, I don't think so because the dev site has the same background movie, inclusive, we have tried to point to the domain (*.azurewebsites.net) that is linked to every app service in azure. This domain opens the same website as the production one, and in this case, Page Speed does not give us any error. It ony shows that error when pointing through www.techtitute.com. We continue investigating any help is appreciated. – David Lorenzo López Jul 09 '20 at 17:15

1 Answers1

0

Finally we have solved the issue by changing traffic manager on Azure from Geographic Routing to Performance Routing.

Now page speed is giving normal scoring

  • So my first instinct was correct then where I said look at the geolocation. 502 is https://dev.techtitute.com/dist/core.js.map, not important but something to sort (and occasionally strange things can happen with PSI / dev tools if they can't access source maps) – GrahamTheDev Jul 10 '20 at 12:50