Huge Pagse 세팅시 확실히 성능 효과가 있네요!! (10% 이상?)
---------------- huge page 안쓸때 ---------------------------------
[enterprisedb@TEST:/ppas/bench]$ sh -x ./a.sh
+ pgbench -n -S -T 30 -c 100 -f test.sql -U ultraman dtest
transaction type: Custom query
scaling factor: 1
query mode: simple
number of clients: 100
number of threads: 1
duration: 30 s
number of transactions actually processed: 184
latency average: 16304.348 ms
tps = 4.517290 (including connections establishing)
tps = 4.561882 (excluding connections establishing)
[enterprisedb@TEST:/ppas/bench]$ sh -x ./a.sh
[enterprisedb@TEST:/ppas/bench]$
+ pgbench -n -S -T 30 -c 100 -f test.sql -U ultraman dtest
transaction type: Custom query
scaling factor: 1
query mode: simple
number of clients: 100
number of threads: 1
duration: 30 s
number of transactions actually processed: 170
latency average: 17647.059 ms
tps = 4.409044 (including connections establishing)
tps = 4.455162 (excluding connections establishing)
---------------- huge page 쓸때 ---------------------------------
[enterprisedb@TEST:/ppas/bench]$ sh -x ./a.sh
+ pgbench -n -S -T 30 -c 100 -f test.sql -U ultraman dtest
transaction type: Custom query
scaling factor: 1
query mode: simple
number of clients: 100
number of threads: 1
duration: 30 s
number of transactions actually processed: 207
latency average: 14492.754 ms
tps = 5.408268 (including connections establishing)
tps = 5.465081 (excluding connections establishing)
[enterprisedb@TEST:/ppas/bench]$ sh -x ./a.sh
+ pgbench -n -S -T 30 -c 100 -f test.sql -U ultraman dtest
transaction type: Custom query
scaling factor: 1
query mode: simple
number of clients: 100
number of threads: 1
duration: 30 s
number of transactions actually processed: 210
latency average: 14285.714 ms
tps = 5.312584 (including connections establishing)
tps = 5.365565 (excluding connections establishing)
[enterprisedb@TEST:/ppas/bench]$
그래서 주인장이 관리하고 있는 ppas 서버에 적용해서 운영하기로 결정!!
