That estimate can be verified with load testing systems like Artillery. My theory is that things would break far sooner than estimated along the following lines:
- Too many WSGI connections if the timeouts aren’t tweaked
- Too many database connections, especially without caching and tuning
- on the Apache side if MaxRequestWorkers isn’t set there will be memory issues with 1GB RAM
- the disk could easily hit IOPS limits, especially if there is a noisy neighbor
It’s not likely all or any of these things will hit IRL, but that all depends on traffic and usage patterns. It matters not, if you were getting 4.2 M requests each day you’d be in the Alexa Top 1000 and could probably shell out for the $8 server :)
- Too many WSGI connections if the timeouts aren’t tweaked
- Too many database connections, especially without caching and tuning
- on the Apache side if MaxRequestWorkers isn’t set there will be memory issues with 1GB RAM
- the disk could easily hit IOPS limits, especially if there is a noisy neighbor
It’s not likely all or any of these things will hit IRL, but that all depends on traffic and usage patterns. It matters not, if you were getting 4.2 M requests each day you’d be in the Alexa Top 1000 and could probably shell out for the $8 server :)