Exact Steps To Migrate ASM Diskgroups To Another SAN Without Downtime.-CSDN博客

本文链接：https://blog.csdn.net/tianlesoftware/article/details/6341946

The present document explains in detail the exact steps to migrate ASM diskgroups from one SAN to another SAN without a downtime.

Solution

If your plans are replacing the current disks associated to your diskgroups with a new storage, this operation can be accomplished without any downtime, so you can follow the next steps

1) Backup all your databases and valid the backup (always required to protect your data).

2) Add the new path (new disks from the new storage) to your asm_disktring to be recognized by ASM:

Example:

SQL> alter system set asm_disktring = '/dev/emcpowerc*' , '/dev/emcpowerh*';

Where: '/dev/emcpowerc*' are the current disks.
Where: '/dev/emcpowerh*' are the new disks.

3) Confirm that the new disks are being detected by ASM:

SQL> select path from v$asm_disk;

4) Add the new disks to your desired diskgroup:

SQL> alter diskgroup <diskgroup name> add disk
‘<new disk 1>’,
‘<new disk 2>’,
‘<new disk 3>’,
‘<new disk 4>’,
.
.
.
‘<new disk N>’ rebalance power <#>;

5) Then wait until the rebalance operation completes:

SQL> select * from v$asm_operation;
SQL> select * from gv$asm_operation;

6) Finally, remove the old disks:

SQL> alter diskgroup <diskgroup name> drop disk
<disk name A>,
<disk name B>,
<disk name D>,
<disk name E>,
.
.
.
<disk name X> rebalance power <#>;

7) Then wait until the rebalance operation completes:

SQL> select * from v$asm_operation;
SQL> select * from gv$asm_operation;

8) Done, your ASM diskgroups and database have been migrated to the new storage.

Note:

Alternatively, we can execute add disk & drop disk statements in one operation, in that way only one rebalance operation will be started as follow:

SQL> alter diskgroup <diskgroup name> add disk '<new device physical name 1>', .., '<new device physical name N>' drop disk <old disk logical name 1>, <old disk logical name 2>, ..,<old disk logical name N> rebalance power <#>;

This is more efficient than separated commands (add disk & drop disk statements).