This content has been marked as final. Show 6 replies
Not sure if now it's certified and worked, but last time I did check, the oracleasm lib didn't worked in OEL6 due SELinux.
oracleasm is released for this version, I installed it using yum and the software repository for this version, so it must be released. It is also working fine, so far. I think the problem is somewhere else, but i still cannot prove this.
SELinux is disabled anyway.
What does the partition table for the CR device look like?
[root@srv01 ~]# fdisk -l /dev/sdf
Disk /dev/sdf: 524 MB, 524288000 bytes
17 heads, 59 sectors/track, 1020 cylinders
Units = cylinders of 1003 * 512 = 513536 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 65536 bytes
Disk identifier: 0x65fc6b88
Device Boot Start End Blocks Id System
/dev/sdf1 1 1020 511500+ 83 Linux
Partition 1 does not start on physical sector boundary.
There are three of those /dev/sdf /dev/sdg /dev/sdh formatted the same way.
first of all you don't need to deinstall everything. Since 18.104.22.168 the root.sh script can be resumed. So simply fix the error and rerun root.sh.
Even if that fails, you still could use $GI_HOME/crs/install/rootcrs.pl -deinstall -force -lastnode to wipe everything. So no need to deinstall everything....
Regarding the error you have -
What does the ASM Log say? Maybe the Diskgroup could not be created correctly (Device permission?), or is already existent?
What more information do you see in the logfile of root.sh (or clusterware logfiles if already existent)?
I want to close the thread with following result:
I went for an installation without ASMLIB. I configured the drives using udev. For some reason the asmlib could not hanldle the permissions on the drives correctly.
After mapping with udev and some reboots to test the availability, I coulnd run the grid infrastructure installation without problems.
Thanks for all the hints.