High Latency in Across EC2 Regions

Hi Bugra,

The resolution depends if you are accelerating your webpage through CloudFront or AWS Global Accelerator, or not.

As a best practice, you would leverage CloudFront to cache your static webpages, providing access to your static contents to everyone around the globe with usually less than 50ms (varies per geographies). That only helps you with static content tho (HTML, CSS, IMG, Videos, etc) which is what you usually measure when we talk about the "page load" time.

For dynamic content, you can also leverage CloudFront or AWS Global Accelerator, which will remove some uncertainty from global routing, as the traffic will get routed to the closest edge location, then AWS will route it through it's backbone to your target instance.

These strategies satisfy the majority of websites to provide the performance needed to their users across the globe (With some help of frontend strategies to minimize the user experience degradation with the dynamic latency, like loading bars and such). If you need < 100ms responses for dynamic content across the globe, that's possible but expensive and technically complex, as you will need to leverage a multi-region approach.

So my advice in a nutshell, accelerate your website with CloudFront, be smart about leveraging it's cache with the proper headers/suffixes, and implement front-end techniques to minimize dynamic latency impact in user experience.