Calculation of the cluster size of the interviewer

Calculation of the cluster size of the interviewer

1. How to determine the cluster size? (Assuming that each server has 8T disk, 128G)

  • There are 100w daily active users every day, and each person averages 100 per day; 100w*100=10000w
  • Each log is about 1k, 100 million per day; 10000 0000 /1024/1024=about 100G
  • Calculate without server expansion within half a year; 100G*180 days = about 18T
  • Save 3 copies; 18T*3 = 54T
  • Reserve 20%~30% Buf = 54T/0.7=77T
  • Go here; about 8T*10 servers

2. How to consider data warehouse layering?
The server is about to expand 1-2 times

Guess you like

Origin blog.csdn.net/qq_42706464/article/details/108651475