数据库性能量化 叶正盛
- 6. 网络性能量化100Mbps/1Gbps/10Gbps带宽:10MB/s, 100MB/s ,1000MB/s本地机房延时:50us-1msmking>ping 10.20.149.82PING 10.20.149.82 (10.20.149.82) 56(84) bytes of data.64 bytes from 10.20.149.82: icmp_seq=0 ttl=64 time=0.124 ms64 bytes from 10.20.149.82: icmp_seq=1 ttl=64 time=0.109 ms64 bytes from 10.20.149.82: icmp_seq=2 ttl=64 time=0.110 ms64 bytes from 10.20.149.82: icmp_seq=3 ttl=64 time=0.109 ms64 bytes from 10.20.149.82: icmp_seq=4 ttl=64 time=0.110 ms
- 15. 分表、分区人员待办工单查询Select * from bpm_work where user_id =‘0001’ and status=‘new’活动数据与历史数据分离:(分表、分区、压缩)工作流(任务流、工单),按状态分表分区历年帐务记录,按年月分表分区
- 25. KV vs RDBMS on SSDKVRDBMSSSDKV数据库与传统数据库对SSD是同等起步,但SSD会让传统数据库满足更多性能需求场景,KV数据库在性能方向优势变小,所以需要在功能、易用性、可维护性方面突破,MongoDB就有它的亮点。