Idrac卡偶尔不可用处理办法

Posted 2020-08-21

R620服务器中，比较经常出现idrac无法连接，或者连接用户数超限的问题，有几个方法可以尝试下：
1、 升级idrac卡的firmware
下载并升级最新的1.57.57版本的firmware：http://www.dell.com/support/home/us/en/04/Drivers/DriversDetails?driverId=XH6FX
该版本的bug fix中提到过一点：
– Fix for issues that cause iDRAC7 sluggish responsiveness after a prolonged period of time (approx. 45-100 days, depending on the usage). In some cases, if the iDRAC is not reset, the iDRAC may become unresponsive and requires a server AC Power on reset. This issue was introduced in firmware release 1.50.50 and fixed in 1.56.55.

该版本解决了idrac启动45-100天后可能出现无法响应的问题。

2、 杀掉超时连接会话
有2个方法：
a. 重启bmc卡
执行命令： ipmitool mc reset cold 即可，将bmc卡重启后，所有的session都会重置释放。

b. 杀掉超时会话
前提是，允许bmc卡通过网络远程连接
ipmitool lan set 1 access on

或者在下面这个地方启用
iDRAC 设置 => 服务 => VNC 服务器 => 超时

或者在idrac卡的下面这个地方设置：
iDRAC 设置 => 网络 => IPMI 设置 => 启用 LAN 上的 IPMI

同时，建议把web server及ssh服务的timeout值适当调低：
iDRAC 设置 => 服务 => Web Server => 超时
iDRAC 设置 => 服务 => SSH => 超时
iDRAC 设置 => 服务 => Telnet => 超时
iDRAC 设置 => 服务 => VNC 服务器 => 超时

利用 racadm 工具（racadm工具的安装自行搞定）关闭超时会话，首先可以查看当前的会话连接情况，例如：
[ 15:41:[email protected]:~ ]#racadm -r 10.59.xx.xx -uroot -p”xx” getssninfo
Security Alert: Certificate is invalid – self signed certificate
Continuing execution. Use -S option for racadm to stop execution on certificate-related errors.
SSNID Type User IP Address Login Date/Time
—————————————————————————
20 SSH root 10.5.xx.xx 11/18/2014 15:38:43
25 GUI root 10.5.xx.xx 11/18/2014 15:40:27
28 RACADM root 10.59.xx.xx 11/18/2014 15:41:17
29 SSH root 10.59.xx.xx 11/18/2014 15:41:18

再执行下面的命令，关闭超时会话连接
[ 15:40:[email protected]:~ ]#racadm -r 10.59.xx.xx -uroot -p”xx” closessn -i 25
Security Alert: Certificate is invalid – self signed certificate
Continuing execution. Use -S option for racadm to stop execution on certificate-related errors.
Session 19 closed successfully.