Hi everyone, I know this isn't stackoverflow but which factors do I keep in mind in scaling compute instances / servers (vertically) to handle more websocket connections?On top of my head I'm thinking:Use cluster to utilize each processorUse flags to increase max memory heap sizeBut what concerns me more is do I scale by CPU count or RAM size? Cloud vendors such as GCP/AWS/AZURE offer different types of compute instances like those with more ram and others with more CPU.Will appreciate any help, thank you.
Submitted July 01, 2018 at 07:23PM by xemasiv
No comments:
Post a Comment