Hi, I have conducted a test on distributed query performance.
I have created 5 server nodes(java) in 3 different server. 40000 records are entered. I created 80 Threads and each thread looped following sql 100 times. select * from user where age<18 (about 50 records is selected) *** age is not key / index After all, I have found that the CPU usage of 2 server nodes (the oldest and the second oldest) jump to a significantly high level (from 5% to 80%-90%) where other 3 nodes increase from 5% to 18% only. Why this happens?