Oracle Exadata 日常维护：服务器下电与加电

Exadata日常维护手册

服务器关机和开机方案

Exadata关机步骤

1. 告警灯检查

下电之前先检查EXdata主机的ib交换机、网管交换机、数据库服务器、存储服务器、PDU有没有黄灯告警

1. 系统健康状态检查
2. infiniband交换机

输入以下命令：

version

ibhosts

showunhealthy

env_test

showpsufru

showtemps

showdisk

dmesg

1. 网管交换机检查

#show running-config

1. 存储服务器检查

dcli –g cell_group –l root “ipmitool sunoem cli ‘show faulty’”

dcli -g cell_group -l root -r "reco" "cellcli -e list physicaldisk"

dcli -g cell_group -l root -r "reco" "cellcli -e list physicaldisk attributes name,errmediacount where disktype=harddisk and errmediacount\>0"

dcli -g cell_group -l root -r "reco" "/opt/MegaRAID/MegaCli/MegaCli64 -Pdlist -a0 |grep 'Slot\|Firmware state'"

dcli -g cell_group -l root "cellcli -e list griddisk attributes name,asmmodestatus,asmdeactivationoutcome"

dcli -g cell_group -l root -r "reco" "biosconfig -get_boot_order | grep DEVICE_NAME"

1. 数据库服务器检查

dcli –g cell_group –l root “ipmitool sunoem cli ‘show faulty’”

dcli -g all_group -l root "/opt/MegaRAID/MegaCli/MegaCli64 -LdInfo -lAll -aAll"

dcli -g all_group -l root "/opt/MegaRAID/MegaCli/MegaCli64 -Pdlist -a0 |grep 'Slot\|Firmware state'"

1. PDU检查

登陆到PDU web页面截图

1. 下电操作
2. 关闭数据库节点的集群和数据库

首先停止所有数据库节点的CRS

使用如下命令：

#dcli - g dbs_group -l root  “/u01/app/11.2.0.3/grid/bin/crsctl stop crs”	

注：如果命令执行失败可在命令行末尾加参数 -f 强制关闭CRS

1. 在命令执行完毕后执行如下命令验证是否所有节点的crs服务都已经停止,如果无输出就说明所有crs都已停止

#dcli -g dbs_group -l root ps -ef |grep d.bin |grep -v grep 

1. 停止存储服务器上的cell服务

#dcli -g cell_group -l root  ”cellcli -e alter cell shutdown services all”

或者

#dcli -g cell_group -l root  /etc/init.d/celld stop 

1. 关闭存储节点的主机

#dcli -g cell_group -l root shutdown -h -y now 

1. 关闭数据库节点的主机

#dcli -g dbs_group shutdown -h -y now

1. 手动关闭机柜PDU-A和PDU-B的开关到OFF状态

Exadata开机步骤

1. 确认所需更换的备件已更换完毕，服务器安放回原机柜位置，各种线缆和原安插位置一致
2. 服务器加电

手动将PDU-A和PDU-B的电源开关调制ON 状态

1. 按启动开关手动启动节点主机
2. 确定infiniband 和cisco交换机完全启动
3. 先给cell节点加电开机，系统正常启动-见图2
4. 然后再给db节点加电开机-见图1

图1 图2

1. 告警灯检查

检查EXdata主机的ib交换机、网管交换机、数据库服务器、存储服务器、PDU有没有黄灯告警

1. 系统健康检查

1.重启完成后，cell服务会自动启动，使用如下命令确认cell服务已正常启动

#dcli -g cell_group -l root /etc/init.d/celld status

2.确认CRS是否正常启动，命令如下：

#dcli -g dbs_group -l root /u01/app/11.2.0.3/grid/bin/crsctl stat res -t

1.infiniband交换机

输入以下命令：

version

nm2version

ibhosts

showunhealthy

env_test

showfruinfo

showpsufru

showtemps

showdisk

showfree

showsmlog

showtopology

1. 网管交换机检查

#show running-config

3.存储服务器检查

dcli -g cell_group -l root -r "reco" "cellcli -e list physicaldisk"

dcli -g cell_group -l root -r "reco" "cellcli -e list physicaldisk attributes name,errmediacount where disktype=harddisk and errmediacount\>0"

dcli -g cell_group -l root "cellcli -e list griddisk attributes name,asmmodestatus,asmdeactivationoutcome,errorcount"

dcli -g cell_group -l root -r "reco" "cellcli -e list griddisk attributes name, asmmodestatus"

dcli -g cell_group -l root -r "reco" "grep MemTotal /proc/meminfo"

dcli -g cell_group -l root -r "reco" "/opt/MegaRAID/MegaCli/MegaCli64 -Pdlist -a0 |grep 'Slot\|Firmware state'"

dcli -g cell_group -l root "cellcli -e list griddisk attributes name,asmmodestatus,asmdeactivationoutcome"

dcli -g cell_group -l root -r "reco" "cellcli -e list physicaldisk attributes name,errmediacount where disktype=harddisk"

dcli -g cell_group -l root -r "reco" "cellcli -e list griddisk attributes name, asmmodestatus,asmdeactivationoutcome,eerorcount "

dcli -g cell_group -l root -r "reco" "biosconfig -get_boot_order | grep DEVICE_NAME"

dcli -g cell_group -l root -r "reco" "biosconfig -get_boot_order | grep DEVICE_NAME"

4.数据库服务器检查

dcli -g all_group -l root imageinfo

dcli -g all_group -l root "ipmitool sunoem cli 'show /SYS product_serial_number'" |grep product_serial_number

dcli -g all_group -l root "/opt/MegaRAID/MegaCli/MegaCli64 -LdInfo -lAll -aAll"

dcli -g all_group -l root "/opt/MegaRAID/MegaCli/MegaCli64 -Pdlist -a0 |grep 'Slot\|Firmware state'"

1. PDU检查

登陆到PDU web页面截图