You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello everyone! How are you? This is my first time commenting on Github, so I apologize in advance if this is not the appropriate topic to comment on here, but I really couldn't find a more convenient place.
Using the docker image from the repository, it is possible to create containers with specific configurations and this is great, I was able to use it without any major problems, however I had the idea of creating services like Refrag that provide on-demand servers for practically 8 dollars per month. This intrigued me because I can't reach this value through AWS Azure and others using VMS or container services without having to manage an instance. Of course, it may not be possible to compete but the values I obtained would be in the range of 54 dollars per month just in cost, without profit, I didn't even include the cost of persistent storage and other things.
Anyway, does anyone have any tips for a viable cloud architecture to use this repository to deliver servers on demand? I tried AWS using ECS + Fargate + EFS to avoid downloading cs2 every time a new container is started. However, this entire architecture is unfeasible.
Again I apologize if this topic is not appropriate here, thank you very much for your attention :)
The text was updated successfully, but these errors were encountered:
Density is king in reducing costs. You will need to manage VMs/Instances and pack them as densely as possible with CS2 Containers.
General Feedback
I use ECS + Fargate + EFS, but only for short lived containers (playing with friends once a week). I find this cost effective.
In EFS, I use storage tiering to put most of the data into the cheap infrequent access tier.
For longer lived servers, this architecture isn't a good fit. Even using EC2 capacity provider for ECS (t3a reserved instances) you probably can't run a server without over-committing for much less than $20/mo.
#24 is interesting because it would remove the need for persistent storage, just:
spin up the container
download a much smaller payload from Steam CDN
apply a customization bundle from a URL if needed
Getting to $8/mo cost
If I were trying to run this as a service for customers, I would:
use ECS with EC2 capacity provider, 1Y reserved instances
ensure customers are aware of a monthly downtime window. I will have Auto-Scaler/Launch Templates perform a rolling replacement of my EC2 instances (cattle, not pets)
Hello everyone! How are you? This is my first time commenting on Github, so I apologize in advance if this is not the appropriate topic to comment on here, but I really couldn't find a more convenient place.
Using the docker image from the repository, it is possible to create containers with specific configurations and this is great, I was able to use it without any major problems, however I had the idea of creating services like Refrag that provide on-demand servers for practically 8 dollars per month. This intrigued me because I can't reach this value through AWS Azure and others using VMS or container services without having to manage an instance. Of course, it may not be possible to compete but the values I obtained would be in the range of 54 dollars per month just in cost, without profit, I didn't even include the cost of persistent storage and other things.
Anyway, does anyone have any tips for a viable cloud architecture to use this repository to deliver servers on demand? I tried AWS using ECS + Fargate + EFS to avoid downloading cs2 every time a new container is started. However, this entire architecture is unfeasible.
Again I apologize if this topic is not appropriate here, thank you very much for your attention :)
The text was updated successfully, but these errors were encountered: