Richmond Database (03)

Saturday 17 May, 2008 - 09:58

The problem described in Richmond Database (02) was overcome by using the procedure in Re-creating ASM Instances and Diskgroups (Metalink DocId: 268481.1).

I then found a problem in that I could not get the ASM instance running on richmond1 because the listener was not active there.  The reason was that the ora.richmond1.vip resource was active on richmond2 .

When I tried the following command, the resource failed to relocate:
crs_relocate ora.richmond2.vip -c richmond1

When I activated debugging on the resource, I found that the default gateway was not set correctly on the eth1 NIC:
crsctl debug log res "ora.richmond2.vip:2"

Once the default route was fixed, the vip resource could then be relocated, the listener started, and the asm instance started:
crs_relocate ora.richmond1.vip -c richmond1
crs_start ora.richmond1.LISTENER_RICHMOND1.lsnr
srvctl start asm -n richmond1

I now have the problem that the disk group DATA will not mount on +ASM2 ( richmond2 ) because there are not sufficient disks available.

/etc/init.d/oracleasm listdisks
ASM1
ASM2
FRA1
FRA2

The disks are visible from there.