[Question] What are the resource requirements of the different configuration modes?

[Breinich]

What are the minimum and preferred resource requirements for the app in different configuration modes?

For example in case of hosting photon too? How does it depend on the regions selected?

[dgraf-gh]

Hi @Breinich,

I think you mean when hosting photon on your own.
In general you can take a look https://github.com/rtuszik/photon-docker?tab=readme-ov-file#available-regions which region are available. I (before I deployed paikka, our own reverse geocoder) had always used photon with the german data set. This is at the moment 7.2GB as the Download. I would calculate with 10 to 12 GB at rest, when you have the update mode parallel enabled, the photon container will download this every 30 days I think. So you would need double the space. Or you go with the planet dataset. That sits at 70GB right now. You can check the sizes here https://download1.graphhopper.com/public/

Please not that only the dbdumps are supported from the photon image.

You could also selfhost nominatim, but there you would need around a 1TB for the whole planet data set.

Back to reitti, you can choose how selfhosted you want to go. For the main usecase, track your movements propably inside one country most of the time. I had photon running with the german dataset. So most of the reverse geocoding would be handled by that photon instance. For which photon did not provide any data, I have geoapify configured. This gives you 3000 request/month free. Which is enough for an occasional holiday outside of germany.

You could also say, i do not want to selfhost photon at all. Then just drop it from the compose file.

I hope that makes it a little bit clearer. Here I have it explained with more detail https://www.dedicatedcode.com/projects/reitti/4.0/configurations/reverse-geocoding/

That is the docs for the upcoming version. So ignore how they are configured, there is a difference between v3 and v4. But the requirements and how it is combined is still valid.