5a8681b5-f0ee-46a2-bed3-62fba619de18

Did you deploy this API in a region near your end user/you?

How are you measuring the latency of the requests?

Yeah Amsterdam and i'm in Belgium

Via the Network tab in DevTools, it's around 40ms to 69ms. Some are 100ms

I got very good ping and internet at my house, so that shouldn't be a big factor.

Are you using the metal edge network? You can check that in your service -> public networking

Yeah

Can you try disabling it and seeing if the latency lowers or is more consistent?

As in downgrading the network?

(clicking that refresh icon?)

If it's not better, can I still go back to Metal edge network?

yes

and yes

It's the same

Maybe i'm used to quicker cuz of local environment and now everything else feels slow

But there probably isn't much I can do right other than improve my NodeJS code for better performance?

details on what the API is doing?

if the API has to handle a lot of work, then here are some solutions that I can think of.

• Do API calls asynchronously.
• Do API calls in big batches all at once
• Do API calls in parallel instead of concurrent.

So basically improve my code

I already do most of those things

ok, so the only thing now is to probably try other regions, or other cloud providers

You could probably to also look into finding out if you could test the latency remotely on a different computer in a different geographical location compared to yours.

So use a service on a different cloud provider, call that api on railway, and log the latency to see if it is a regional thing.

this is likely the case. api calls to a local server will have 0ms rtt, plus the actual processing that the api is doing

more details on what processing is actually being done by your api would be helpful for suggesting optimizations

40ms is not bad

Oh okay 🙂

Thanks guys