On the other hand, if I look at stats, apparently thanks to ARGO pages load faster (see screenshot), but not always. Sometimes, it looks like data loads faster without ARGO.
The quality of your content has decreased? User interest in your topics has decreased? Correlation and causation aren’t the same. From a performance standpoint the website is faster. With that data alone user experience has improved. External factors are just that.
This delta, especially over a rather short time, also tends to be on the statistical noise level.
If it persists over weeks, and you can exclude the root causes cscharff put forth, you should investigate again.
In the meantime there may be one possible reason, and that is an undiscovered increase in error rate. But you’d have to peruse your own logs (main site and CF) and/or engage in - likely expensive - external distributed testing service to find out if that even is the case.
It would be helpful if we had the option to evaluate Argo performance over a time period longer than 48h, and if we had the median instead of the mean as a metric.
El dv., 9 de des. 2022, 9:56, Stephan-X via Cloudflare Community <[email protected]> va escriure: