Environment
PlateSpin Protect 10.3
Situation
This article describes how to workaround to resolve the below error during failback of SLES version 9 workload.
could not add partition /dev/sda1 to monitoring list: Unable to find mountpoint for "/dev/sda1" PATH is :/lib/modules/2.6.5-7.97-default/platespin/blkwatch.ko using /dev/blkwatch_1355 device, errno: 0
Resolution
1. Manually modify /etc/fstab to use device-name instead of UUID.
For Example:
A sample of /etc/fstab with UUID:
UUID=b37fa7c6-ad26-4b65-8760-6467701e7e79 / reiserfs acl,user_xattr 1 1
Assuming its known that UUID b37fa7c6-ad26-4b65-8760-6467701e7e79 relates to /dev/sda1, then the line above should be changed to as follows:
Assuming its known that UUID b37fa7c6-ad26-4b65-8760-6467701e7e79 relates to /dev/sda1, then the line above should be changed to as follows:
/dev/sda1 / reiserfs acl,user_xattr 1 1
To view the relation between device name and UUID run the blkid command. Below is a sample output:
To view the relation between device name and UUID run the blkid command. Below is a sample output:
# blkid
/dev/sda1: UUID="b37fa7c6-ad26-4b65-8760-6467701e7e79" TYPE="reiserfs"
/dev/sda2: TYPE="swap"
/dev/sda1: UUID="b37fa7c6-ad26-4b65-8760-6467701e7e79" TYPE="reiserfs"
/dev/sda2: TYPE="swap"
2. Remount all volumes involved or reboot the source workload if the remount command throws an error as volumes are busy.
3. Remove the workload and readd from scratch.
3. Remove the workload and readd from scratch.
Cause
SLES version 9 has the root ("/") not listed in "/proc/mounts". This causes the lookup (Device-Name => MountPoint) to return null value, hence operation fails.