linux networkmanager 服务,Linux,Network manager 导致节点异常重启

推断是Network manager 导致的，原因待查

今天在VmWare的虚拟机上装了个测试RAC，又遇到了一个摸不到头绪的问题

CRS装好后，一旦登陆图形界面，节点就重启，事情就有这么巧

不登陆图形界面，观察了1个小时没问题，一旦登陆后，立刻重启

在OS日志中，一旦登陆图形界面，重启前的日志如下

Sep  5 19:29:18 dm01db01 nm-system-settings: Loaded plugin ifcfg-rh: (c) 2007 - 2008 Red Hat, Inc.  To report bugs please use the NetworkManager mailing list.

Sep  5 19:29:18 dm01db01 nm-system-settings:    ifcfg-rh: parsing /etc/sysconfig/network-scripts/ifcfg-lo ...

Sep  5 19:29:18 dm01db01 nm-system-settings:    ifcfg-rh: parsing /etc/sysconfig/network-scripts/ifcfg-eth1 ...

Sep  5 19:29:18 dm01db01 nm-system-settings:    ifcfg-rh:     read connection 'System eth1'

Sep  5 19:29:18 dm01db01 nm-system-settings:    ifcfg-rh: parsing /etc/sysconfig/network-scripts/ifcfg-eth0 ...

Sep  5 19:29:18 dm01db01 nm-system-settings:    ifcfg-rh:     read connection 'System eth0'

集群件没有任何日志，就像机器被人直接重启了一样，找不到任何原因

ping心跳，偶尔有200多ms，但是重启前，ping都在几ms内

vmstat监控，CPU利用率也没有问题

测试了如下调整：

1.加大 misscount 无效

2.调整 diagwait，也没有任何日志

3.关闭了无用的服务，无效

4.重新换了个网段，无效

一直觉得是网络的问题，搜索关键字 ifcfg-rh ，找到了一篇文章 OEL: Error: Missing Or Invalid IP4 Prefix '0' On Linux Server (Doc ID 1522095.1)

虽然现象和我的问题无关，但是抱着死马当活马医的想法，跟着文档关闭了Network manager

1.在/etc/sysconfig/network-scripts/ifcfg-eth* 中增加 NM_CONTROLLED="no"

2.chkconfig NetworkManager off

3.reboot

重启后主机正常。在OS日志中看到：

Sep  5 19:41:06 dm01db01 nm-system-settings: Loaded plugin ifcfg-rh: (c) 2007 - 2008 Red Hat, Inc.  To report bugs please use the NetworkManager mailing list.

Sep  5 19:41:06 dm01db01 nm-system-settings:    ifcfg-rh: parsing /etc/sysconfig/network-scripts/ifcfg-lo ...

Sep  5 19:41:06 dm01db01 nm-system-settings:    ifcfg-rh: parsing /etc/sysconfig/network-scripts/ifcfg-eth1 ...

Sep  5 19:41:06 dm01db01 nm-system-settings:    ifcfg-rh:     read connection 'System eth1'

Sep  5 19:41:06 dm01db01 nm-system-settings:    ifcfg-rh: Ignoring connection 'System eth1' and its device because NM_CONTROLLED was false.

Sep  5 19:41:06 dm01db01 nm-system-settings:    ifcfg-rh: parsing /etc/sysconfig/network-scripts/ifcfg-eth0 ...

Sep  5 19:41:06 dm01db01 nm-system-settings:    ifcfg-rh:     read connection 'System eth0'

Sep  5 19:41:06 dm01db01 nm-system-settings:    ifcfg-rh: Ignoring connection 'System eth0' and its device because NM_CONTROLLED was false.

可以看到配置被忽略掉了。

先记录一个，以后在研究

版本信息

[root@dm01db01 network-scripts]# cat /etc/issue

Oracle Linux Server release 5.9

Kernel \r on an \m

[root@dm01db01 network-scripts]# uname -a

Linux dm01db01 2.6.39-300.26.1.el5uek #1 SMP Thu Jan 3 18:31:38 PST 2013 x86_64 x86_64 x86_64 GNU/Linux

[root@dm01db01 network-scripts]# /u01/app/oracle/product/crs/bin/crsctl query crs activeversion

CRS active version on the cluster is [10.2.0.5.0]