Implement ASM failure groups

 

 

Recovering disk failure.

After failure of Esx1 server, the cluster is still UP but with some missing disk…

Only two voting files.

-bash-4.1$ crsctl query css votedisk
## STATE File Universal Id File Name Disk group
— —– —————– ——— ———
1. ONLINE 7febd2c7aefd4fb9bf00c0712aa85577 (/dev/asmGRID2) [GRID]
2. ONLINE 385705904fe84f04bf9f82e05e8830a0 (/dev/asmGRID3) [GRID]
Located 2 voting disk(s).
-bash-4.1$

select name ,FAILGROUP,MOUNT_STATUS,MODE_STATUS,STATE from v$asm_disk;

NAME FAILGROUP MOUNT_S MODE_ST STATE
———————– ——- ———————- ——- ——-
CLOSED ONLINE NORMAL
CLOSED ONLINE NORMAL
CLOSED ONLINE NORMAL
CLOSED ONLINE NORMAL
_DROPPED_0001_DATA DATA_0001 MISSING OFFLINE FORCING
_DROPPED_0000_RECO RECO_0000 MISSING OFFLINE FORCING
_DROPPED_0000_FRA FRA_0000 MISSING OFFLINE FORCING
FRA_0001 FRA_0001 CACHED ONLINE NORMAL
GRID_0002 GRID_0002 CACHED ONLINE NORMAL
RECO_0001 RECO_0001 CACHED ONLINE NORMAL
GRID_0001 GRID_0001 CACHED ONLINE NORMAL
DATA_0000 DATA_0000 CACHED ONLINE NORMAL

At this stage, it’s no more possible to put back missing disk online, event if they are connected.

SQL> alter diskgroup data drop disk DATA_0001;
alter diskgroup data drop disk DATA_0001
*
ERROR at line 1:
ORA-15032: not all alterations performed
ORA-15054: disk “DATA_0001” does not exist in diskgroup “DATA”

To add back missing disk, we have to alter diskgroups and add the missing disk.

Let’s start with the GRID diskgroup:

SQL> alter diskgroup grid add disk ‘/dev/asmGRID1’;
alter diskgroup grid add disk ‘/dev/asmGRID1’
*
ERROR at line 1:
ORA-15032: not all alterations performed
ORA-15033: disk ‘/dev/asmGRID1’ belongs to diskgroup “GRID”

Oups, the disk is detected being part of GRID diskgroup…

To clean it us the following as root.

[root@node2 ~]# dd if=/dev/zero bs=8k count=10000 of=/dev/asmGRID1

SQL> alter diskgroup grid add disk ‘/dev/asmGRID1’;

Diskgroup altered.

SQL>

-bash-4.1$ crsctl query css votedisk
## STATE File Universal Id File Name Disk group
— —– —————– ——— ———
1. ONLINE 7febd2c7aefd4fb9bf00c0712aa85577 (/dev/asmGRID2) [GRID]
2. ONLINE 385705904fe84f04bf9f82e05e8830a0 (/dev/asmGRID3) [GRID]
3. ONLINE 7fff38e489ab4f99bfe06ab045ff968c (/dev/asmGRID1) [GRID]
Located 3 voting disk(s).

Same has to be done for DATA, FRA and RECO diskgroups…

 

The following two tabs change content below.

Jacques

I am Oracle Certified Master 11g & 12c database architect with significant experience in heterogeneous environments, and strong ability to lead complex and critical projects requiring multiple technical implementations. at Trivadis SA

1 thought on “Implement ASM failure groups

Leave a Reply

Your email address will not be published. Required fields are marked *