Performance de JavaServer Faces 1.2 no GlassFish
Achei um artigo interessante na rede relatando um estudo de performance do JSF R.I. 1.2 juntamente com o GlassFish v1 e v2 , usando o modo de client e server para salvar a informacao de estado dos objetos:
1) State saving method = server
1) State saving method = server
| |
1-CPU |
2-CPUs |
4-CPUs |
8-CPUs |
10-CPUs |
| SJSAS 9.0 PE FCS JSF 1.2 FCS |
metric= 57.74 totalOps= 17321 |
metric= 85.18 totalOps= 25555 |
metric=177.25 totalOps=53176 |
metric=190.52 totalOps=57155 |
metric=166.16 totalOps=49847 |
| SJSAS 9.1 PE b27 JSF 1.2_03 RC3 |
metric= 70.80 totalOps= 21239 |
metric= 129.86 totalOps= 38957 |
metric=321.36 totalOps=96409 |
metric=458.23 totalOps=137469 |
metric=557.30 totalOps=167189 |
2) State saving method = client
| |
1-CPU |
2-CPUs |
4-CPUs |
8-CPUs |
10-CPUs |
| SJSAS 9.0 PE FCS JSF 1.2 FCS |
metric= 27.03 totalOps= 8108 |
metric= 36.12 totalOps= 10836 |
metric=67.34 totalOps=20202 |
metric=59.01 totalOps=17704 |
metric=55.84 totalOps=16752 |
| SJSAS 9.1 PE b27 JSF 1.2_03 RC3 |
metric= 43.39 totalOps= 13017 |
metric= 72.97 totalOps= 21892 |
metric=174.86 totalOps=52457 |
metric=296.68 totalOps=89003 |
metric=335.02 totalOps=100505 |
number of users = number of CPUs
metric = number of operations/transactions per second
application = 1.1 based jsf-cardemo
Publicado por Dyego Souza do Carmo em 23 de Janeiro de 2007 10h56min43s BRST
#