Why we built an Edge benchmarking tool for Turso

#Why we did it

“Yeah but how much latency can it really save me?”

We often get questions like this, and it's hard to properly convey in writing the potentially drastic latency differences between a typical centralized cloud database scenario and what Turso offers at the Edge. Even infographics can be read as potentially unrealistic or outright inaccurate.

So we built a simple benchmarking tool using real, live latency data to make it simple to visualize this in an interactive way.

#What it does

The tool offers a simple toggle that showcases real latencies to a number of simulated fairly typical user locations around the world from a central cloud database location in North America and, when the toggle is switched, shows real latencies to those same simulated user locations from Turso Edge database replicas placed closer to them.

In most cases, the latency differential is stark and immediately noticeable.

#How we did it

We used live data from Planetfall to represent real, current latencies around the world from simulated users to real world cloud regions you may be familiar with. Then, we used the same simulated user locations to display real, current latencies from real Turso edge locations closer to those users where the product allows for easy local replicas of a database instance. Planetfall measures the latency to the database every 60s. We then compute the 5-minute moving average of these points, and refresh the UI every 6 seconds.

Because Turso makes it fast and cost effective to replicate your database closer to your users, and because modern user bases are often distributed in various clusters around the world, even for midsized apps and services, the Edge database toggle in the tool represents real world conditions and not some unrealistic, idealized scenario.

#Check it out

If you haven't already, we recommend you check out the tool at turso.tech/edge and, of course, you can sign up for Turso for free and try it out yourself.