Ok so i want to know the pricing of deployments. I am trying to make a really small api for transfering and holding saves for a game, and im wondering what the pricing is and what the best deployment option is
You can find the prices for the deployment on the prices page.
Prices for different types of deployments:
Autoscale: starting at $0.000024/second (you only pay for the seconds of deployment operation when someone uses it)
Reserved VM: starting at $7/month (it does not depend on whether the deployment was used or not)
Static: free (for free up to a certain limit, if you exceed the limit, see the documentation for more information)
It’s hard to say which deployment is the best. Each type of deployment is designed for different projects (it’s like asking “which is better, a hammer or a screwdriver”).
The purposes of different types of deployments:
Autoscale: it is used to launch sites with a backend (Flask, Django, etc. frameworks can be used), for more information; see the documentation
Reserved VM: it is necessary for those replays that should always work without stopping, for example, for bots (unlike Autoscale, which goes into “sleep” mode when it is not used much); see the documentation
I have a specific use case of where I want to deploy python code for an AI bot and I’m not sure if I should chose Auto Scale or Reserved VM. And I want to know how many requests can a Reserved VM can handle at a time.
For bots, you should use a Reserved VM deployment. Autoscale is used for websites.
And how many requests can a Reserved VM can handle at a time?
As far as I know, when deploying, you can specify the CPU power and RAM size. That is, you can adjust the power of your Reserved VM and the number of requests received.
But there is nothing that shows that how many requests this specific server can handle.
It seems to me that either there is no exact number, or this number is very large, so you should not worry about it.
ok, but there are multiple options in Reserved VM so there must be some difference in their capabilities can you provide any details on it, you guys must mention some criteria to differentiate between their capabilities.
Do you mean Web server or background worker?
If you run the Reserved VM as a web server, the url you specified will be listened to, that is, it will work as Autoscale, but without shutting down. This method is suitable for APIs that should be running forever.
If you run the Reserved VM as a background worker, then the url will not be listened to, that is, the code will simply not turn off. This method is suitable for launching bots.