Node Eviction due to OLOGGERD High CPU
来源:互联网 发布:知之一字 众妙之门 编辑:程序博客网 时间:2024/06/11 02:55
1、从网上下载了iopp.c,编译 gcc -o iopp iopp.c,
运行该命令iopp -c -i -k -u 2
发现ologgerd进行磁盘IO很高,该进程是CHM(Cluster Health Monitor)导致的,CHM是Oracle11g的新特性,主要用来收集节点上的负载信息,网上说是bug,再没有更好的处理方式下先关闭。
2、关闭CHM
Node Eviction due to OLOGGERD High CPU (Doc ID 1636942.1)
Loggerd uses high cpu and do lots of I/O to the disk where the BDB (Berkeley Database used by CHM) resides.
This is due to BUG 13867435 - OLOGGERD USING A LOT OF RESOURCES .
你把CHM关闭看看如何
[root@node3 ~]# ps -ef |grep crf
root 3127 1 1 22:10 ? 00:00:02 /u01/11.2.0/ghome/bin/ologgerd -M -d /u01/11.2.0/ghome/crf/db/node3
root 3488 3264 0 22:12 pts/0 00:00:00 grep crf
[root@node3 ~]# cd /u01/11.2.0/ghome/bin/
./crsctl stop res ora.crf -init
CRS-2673: Attempting to stop 'ora.crf' on 'node3'
CRS-2677: Stop of 'ora.crf' on 'node3' succeeded
[root@node3 bin]# ps -ef |grep crf
root 3609 3264 0 22:13 pts/0 00:00:00 grep crf
[root@node3 bin]#
关闭后磁盘Io降下来了,性能恢复正常。
感悟:通过这件事以后我会尝试发现问题解决问题
以前看到磁盘爆满就想当然的以为是磁盘文题,自己认为是共享磁盘不断通信造成的搞IO
想到了用另一种方法重建asm,比如openfile。还想到了换SSD
这次因为前几天都是ok的,当然也有幸运成分,肯定是前几天没触发CHM进程
这次我就试着监控linux系统,然后试着找占用磁盘高的进程,然后再从rac方面找
也是幸运的我看到了这一篇文章,其实my support上应该也可以找到
运行该命令iopp -c -i -k -u 2
发现ologgerd进行磁盘IO很高,该进程是CHM(Cluster Health Monitor)导致的,CHM是Oracle11g的新特性,主要用来收集节点上的负载信息,网上说是bug,再没有更好的处理方式下先关闭。
2、关闭CHM
Node Eviction due to OLOGGERD High CPU (Doc ID 1636942.1)
Loggerd uses high cpu and do lots of I/O to the disk where the BDB (Berkeley Database used by CHM) resides.
This is due to BUG 13867435 - OLOGGERD USING A LOT OF RESOURCES .
你把CHM关闭看看如何
[root@node3 ~]# ps -ef |grep crf
root 3127 1 1 22:10 ? 00:00:02 /u01/11.2.0/ghome/bin/ologgerd -M -d /u01/11.2.0/ghome/crf/db/node3
root 3488 3264 0 22:12 pts/0 00:00:00 grep crf
[root@node3 ~]# cd /u01/11.2.0/ghome/bin/
./crsctl stop res ora.crf -init
CRS-2673: Attempting to stop 'ora.crf' on 'node3'
CRS-2677: Stop of 'ora.crf' on 'node3' succeeded
[root@node3 bin]# ps -ef |grep crf
root 3609 3264 0 22:13 pts/0 00:00:00 grep crf
[root@node3 bin]#
关闭后磁盘Io降下来了,性能恢复正常。
感悟:通过这件事以后我会尝试发现问题解决问题
以前看到磁盘爆满就想当然的以为是磁盘文题,自己认为是共享磁盘不断通信造成的搞IO
想到了用另一种方法重建asm,比如openfile。还想到了换SSD
这次因为前几天都是ok的,当然也有幸运成分,肯定是前几天没触发CHM进程
这次我就试着监控linux系统,然后试着找占用磁盘高的进程,然后再从rac方面找
也是幸运的我看到了这一篇文章,其实my support上应该也可以找到
0 0
- Node Eviction due to OLOGGERD High CPU
- Reducing CPU problem caused due to high buffer gets (after large delete)
- Node reboot or eviction: How to check if your private interconnect CRS can transmit network heartbea
- High Waits on 'Db File Sequential Read' Due to Table Lookup Following Index Access
- ORACLE RAC节点意外重启Node Eviction诊断流程图
- bug宝典Node篇 Fail to decode request due to: RpcInvocation
- 网站High CPU分析
- High Waits on 'Db File Sequential Read' Due to Table Lookup Following Index Access (文档 ID 875472.1)
- simulation is not running in real time due to excessive cpu load
- Agile Plm Notifications Are Not Being Sent Due to Node 0 Error
- CRS-0184 and clssnmLocalJoinEvent: takeover aborted due to ALIVE node on Disk
- Database can't be started on Node 1 due to serverpool Max value is 1
- Access restriction: The type Node is not accessible due to restriction on required library C: \Prog
- due to, proportion.
- portscan due to nmap
- flume kafka-sink high cpu
- High CPU Dump收集工具
- Oracle Grid Infrastructure: Understanding Split-Brain Node Eviction (Doc ID 1546004.1)
- 开源项目Make Point
- 不使用quartz实现定时,使用@Scheduled注解实现定时--一次配置到处使用
- Nutz中过滤特殊字符
- nginx 或tengine 访问日志分割处理
- ios uiwindow 的用法
- Node Eviction due to OLOGGERD High CPU
- 试题01 A+B和C
- ios 9 隐藏 StatusBar
- 下拉放大headView
- 四道Java基础题 你能对几道?
- 常用sqlite的命令收集贴
- Qt之界面实现技巧
- [leetcode] 40. Combination Sum II 解题报告
- Imagemagic的安装