log file sync事件Avg wait (ms)时间长
我的log file sync事件Avg wait (ms)时间长,如下所示:(log file sync平均等待时间10ms)Top 5 Timed Foreground Events
Event Waits Time(s) Avg wait (ms) % DB time Wait Class
log file sync 266,754 2,706 10 45.88 Commit
DB CPU 2,180 36.95
log file parallel write的平均等待时间也是10ms
Background Wait Events
Event Waits %Time -outs Total Wait Time (s) Avg wait (ms) Waits /txn % bg time
log file parallel write 280,824 0 2,680 10 1.11 62.63
我想着应该是物理存储设备IO造成的吧,两个共享存储间做了镜像,一个复制到另外一个。附件是awr报告,请帮忙看下。 我想应该是io问题,用raccheck跑分,发现redo log写慢,如下:
WARNING OS Check Redo log write time is more than 500 milliseconds 基本信息:安装的oracle 11.2.0.3 rac环境 2节点
操作系统:Red Hat Enterprise Linux Server release 5.8 (Tikanga)
Avg global cache current block pin time (ms): 118,436.9
Avg global cache current block flush time (ms): 1,898,667.9
这两个指标很诡异~~,私网连接是怎么设置的?
还有一个user rollbacks达到每秒22次
而user commits只有每秒1.4次。 为啥有这么多的user rollbacks? 能否减少user rollback的次数? 粗略看了一下,大致可能对 t_aaa_failure_record 表相关的两三个sql(应该是全表扫描)优化一下,
系统的IO,及节点的通信可以减小一半 ~ 每个节点 减少之间的通信开销
让应用固定在某个节点上运行 11g 可以在空闲时 用11g新特性之IO校准(IO Calibration) 测一下
http://www.askmaclean.com/archives/11g%E6%96%B0%E7%89%B9%E6%80%A7%E4%B9%8Bio%E6%A0%A1%E5%87%86io-calibration.html
页:
[1]