Forums

Part of a little idea for a project I wanted to experiment with, contains logging some real-time data (about 5 kB) approximately every 30 seconds. This data would be sent to the python web app on PythonAnywhere via a local client script running on my laptop.

So I tried reading up on limitations to find out if the whole thing is viable. A beginner's web app is limited to 1 web worker, which doesn't seem to matter in this case, as there is only that 1 connection for a short time every 30 seconds. However, I read on bandwidth limits in Why the fuzzy bandwidth limits?, but still don't really know how to gauge "low bandwidth". I couldn't find anything about frequency of queries to the web app? I don't want my requests to break the limits or rules or even appear as some kind of unintentional DDoS attack slowing everyone down. :-/

Summing up, for a full day the web app would yield

• ~15 MB of traffic
• ~3000 queries

Since I'm pretty new to web app development, I have no idea: is that beyond the realms of fair-use - or more like boring, standard numbers for a small project? I hope it's alright to post this very specific inquiry. I'm sorry, if I did miss something.

Thanks already :-)

Specific queries are great, they're the easiest to answer ;-)

3,000 queries a day with about 5kB coming in should be absolutely fine, so long as you're not doing any heavy processing in the web app views. So, for example, if you're simply putting them in a database, that's no problem at all.

The reason we're not more specific about the number of requests is because web apps vary wildly in the amount of processing each request needs. So, reading a file and sending back the contents, or reading/writing something simple from a database is practically free from a processing perspective, so it's entirely welcome.

At the far opposite extreme... once we found someone who was trying to make their web request processors mine for bitcoins, which (a) was silly because no matter how much free processing they got from us, the chance of them ever making any money with CPU mining was basically zero, and (b) was hogging a huge amount of processing power. That definitely wasn't OK.