2 months ago
Hi,
I have a $20 pro plan and it says 32 CPU cores and 32 GB RAM for every deployed service. However on the dashboard the overall services doesn't event consume 1 cpu and 1GB RAM of all teh services combined. The advertised vs in practice is no match.
Can someone please clarify how are these get allocated in practice. I found the applications run very slow and no way I could bind these dedicated to some of the services where they are required.
Thanks,
Naran
1 Replies
2 months ago
This thread has been marked as public for community involvement, as it does not contain any sensitive or personal information. Any further activity in this thread will be visible to everyone.
Status changed to Open Railway • about 2 months ago
2 months ago
If you application isn't doing very much, it will not use much CPU time or RAM. Assigning more to a service just means it can use more CPU time/RAM if it needs to.
It doesn't inherently make things faster unless your application can make use of multiple threads, or needs more RAM.
As for your application feeling slow, this is likely down to the vCPU architecture. Your application runs in a container on a shared host, so other applications will be using up "real" CPU time, making them less responsive than if they were running on a dedicated server - you will see this usage if you run something like htop via SSH in one of your containers as it reports the host CPU usage, not the container's.
Railway offers dedicated hosts as part of their $10,000/mo Committed Spend tier if your performance needs are mission-critical.