Oracle Enterprise Linux 5.8: Multipathed disk partitions are not discovered properly

I just came accross a situation where multipathed devices were not discovered correctly. The device itself was discovered but the disks partitions were missing.

It turned out there are two bugs in OEL 5.8. The first one is related to multipath itself and can be fixed like this:

ls -l /sbin/mu*
 -rwxr-xr-x 1 root root     0 Jun 20 11:01 /sbin/multipath
 -rwxr-xr-x 1 root root 75464 Jul  9  2011 /sbin/multipathd
ln -s /sbin/multipath /sbin/multipath.static
ls -l /sbin/mu*
 -rwxr-xr-x 1 root root     0 Jun 20 11:01 /sbin/multipath
 -rwxr-xr-x 1 root root 75464 Jul  9  2011 /sbin/multipathd
 lrwxrwxrwx 1 root root    15 Jun 20 11:02 /sbin/multipath.static -> /sbin/multipath

The second issue arises with a huge number of LUNs (usually > 100). Here the HAL daemon sometimes has not enough time to detect all the disks. As a workaround create the following file with the following contents:

 cat /etc/sysconfig/haldaemon
 --child-timeout=720

Once done reboot the server and you are done.

This entry was posted in Oracle in general. Bookmark the permalink.

Leave a Reply

Your email address will not be published.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