[MVS]Oracle Exadata X8存储主动更换硬盘

硬盘在系统日志提示有error，但ASM多次尝试修复，影响数据库稳定，主动更换硬盘。

查看存储节点系统日志，多次提示硬盘error，但ASM接管重新挂载硬盘后，恢复正常。

Oracle Exadata X8存储主动更换硬盘

环境：Oracle Exadata X8

exadata system info: 19.2.13.0.0.200428

OS: oracle linux 7.7

硬盘在系统日志提示有error，但ASM多次尝试修复，影响数据库稳定，主动更换硬盘。

1.检查硬盘号；

查看硬盘插槽，硬盘上承载的grid disk等；

# cellcli list diskmap

或：# cellcli list physicaldisk   

或  # cellcli –e "list diskmap" | grep 'X:Y'

2.删除硬盘

2.1. 如Oracle Exadata System Software release 21.2.0,则如下命令将自动删除硬盘并管理冗余性；

CellCLI> alter physicaldisk X:Y drop for replacement maintain redundancy

注意：等命令执行结束才能下一步；

2.2：如Oracle Exadata System Software release 小于21.2.0，则需要顺序执行如下命令：

1）、.删除对应grid disk;

SQL> ALTER DISKGROUP diskgroup_name DROP DISK asm_disk_name;

备注：如删除griddisk报错

ORA-15032: not all alterations performed

ORA-15410: Disks in disk group OCRVT do not have equal size.

或者

ORA-15032: not all alterations performed

ORA-15411: Failure groups in disk group OCRVT have different number of disks.

解决方案：

动态修改这两个隐藏参数解决限制副本数问题：

alter system set "_asm_disable_failgroup_size_checking"=true;

alter system set "_asm_disable_dangerous_failgroup_checking"=true;

2) . 等待磁盘rebalance操作完成，查询：select  * from v$asm_operation; 

3). cell中删除物理硬盘

CellCLI> alter physicaldisk X:Y drop for replacement

如防止超时退出，可后台执行：# cellcli -e "alter physicaldisk X:Y drop for replacement" & 

备注：

• ϒ⁄ cell alert log  cell中查看进度， 日志路径：/opt/oracle/cell/log/diag/asm/cell/<cell_name>/trace/alert.log
• ϒ⁄ 查看总体脏块刷新速度，最终是否结束以日志提示为准。CellCLI> list metriccurrent attributes name,metricvalue where name like 'FC_BY_DIRTY';  
• ϒ⁄ exadata 19.2版本可能会hang,  但本次测试解决方案效果不明显；Exadata ALTER PHYSICALDISK N:N DROP FOR REPLACEMENT is hung (Doc ID 2574663.1)   
• 14T HDD硬盘删除耗时5小时左右；

3.Ensure that the blue OK to Remove LED on the disk is lit before removing the disk.

硬盘删除成功后，# cellcli list physicaldisk 查看磁盘已删除；

4.Replace the new hard disk.

手动更换硬盘后，系统自动按原配置加到ASM 磁盘组中；

5.verify the LUN, cell disk and grid disk associated with the hard disk were created.

CellCLI> alter cell validate configuration     》》》自动升级固件；不能有mdadm错误； 

CellCLI> list lun lun_name

CellCLI> list celldisk where lun=lun_name

CellCLI> list griddisk where celldisk=celldisk_name

CellCLI> LIST PHYSICALDISK  detail

6.硬盘更换完成后将ASM隐藏参数修改回来；

alter system set "_asm_disable_failgroup_size_checking"=false;

alter system set "_asm_disable_dangerous_failgroup_checking"=false;

7.Verify the grid disk was added to the Oracle ASM disk groups.

The following query should return no rows.

SQL> SELECT path,header_status FROM v$asm_disk WHERE group_number=0;

The following query shows whether all the failure groups have the same number of disks:

SQL> select * from v$asm_disk;

参考：

官方文档：3.3.5 Replacing a Hard Disk Proactively

https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmn/replacing-hard-disk-proactively.html

 

Oracle Exadata换盘操作-Replacing a Hard Disk Proactively

https://blog.csdn.net/xxzhaobb/article/details/126389951

 

Exadata存储节点，主动更换磁盘

https://www.cnblogs.com/missyou-shiyh/p/18581627

 

MOS:

How to Replace a Hard Drive in an Exadata Storage Cell Server (Predictive Failure) (Doc ID 1390836)