Skip to main content
ONTAP FLI

ONTAP FLI Linux host remediation

Contributors netapp-barbe netapp-aherbin

Before running the ONTAP FLI migration on a Linux host, complete these pre-migration steps. There are separate steps for offline and online migrations.

Pre-Migration - Offline FLI

These steps are common for both online and offline FLI. Before unprovisioning the source LUN from the host, ensure the following steps are completed.

Filesystem (LVM & Non-LVM) Pacemaker with LVM (SCSI3 PR) SANBOOT

Step 1: Collect end-to-end Block Device configuration: multipath health/paths, LVM layout on the block device and GRUB/bootloader settings confirming the correct boot device and boot parameters.

Step 1: Collect end-to-end Block Device configuration: multipath health/paths, LVM layout on the block device and GRUB/bootloader settings confirming the correct boot device and boot parameters.

Step 1: Collect end-to-end SAN-boot configuration: multipath health/paths, LVM layout on the boot LUN, and GRUB/bootloader settings confirming the correct boot device and boot parameters.

Step 2: Make sure the /etc/fstab entries for the respective data devices (LVM/Non-LVM) are referenced by their FS UUID.

Step 2: Disable the cluster resource groups for the respective LUNs being migrated. Once disabled, delete the resource groups.

Step 2: Make sure the /etc/fstab entries for SAN Boot device including the data devices (lvm/nonlvm) are referenced by their FS UUID.

Step 3: Ensure all the operations to the target mount points are paused. Ensure there are no processes still accessing the block device. If required, drain/kill the open process handles.

Step 3: Perform the LVM VG configuration backup to a file for the respective resource groups that we just disabled/deleted. Followed by Deactivation of the VGs.

Step 3: Ensure that you have updated the GRUB_CMDLINE_LINUX line with the correct UUID references. Recreate a new GRUB settings based on the UUID device references.

Step 4: Umount the filesystem and perform the LVM volume group configuration backup. Deactivate the VG and export the VG

Step 4: Remove the source LUNs from the PCS stonith configuration.

Step 4: Validate the changes with the new grub settings and recreate the new initrd image to reflect changes with SAN Boot Device.

Step 5: Flush the multipath configuration for the device and remove all the stale SCSI device files.

Step 5: Flush the multipath configuration for the device and remove all the stale SCSI device files.

NOTE: If any lun persistent reservations is seen, clear them and ensure no PR keys are

Step 5: Reboot the host to ensure the host can come up fine the above changes. Once the host comes up fine, understand the host pre remediation for FLI migration is ready.

Step 6: Reload the multipath configuration so that the host has the updated device list. If required, perform the host SCSI Bus rescan.

Step 6: Rescan the host to ensure there are no stale entries in the existing device mapping.

Step 6: Now shutdown the sanboot host, to proceed with ONTAP FLI workflow.

Pre-Migration - Online FLI

Below are specific steps of online FLI but these steps must be preceded by steps in the pre-migration offline FLI. The following steps must be carried out after the NetApp LUN is provisioned to the host. These steps must be carried out after the NetApp LUN is provisioned to the host. And it is applicable only the online FLI.

Filesystem (LVM & Non-LVM) Pacemaker with LVM (SCSI3 PR) SANBOOT

Step 1: Rescan storage on all the mapped Linux hosts.

Step 1: Rescan storage on all the mapped Linux hosts.

Step 1: PowerON the server.

Step 2: Discover the multipath devices.

Step 2: Discover the multipath devices. Ensure the migrated LUNs are added to PCS Stonith Configuration.

Step 2: On the server HBA BIOS or UEFI configure the new boot device that is presented to this Host HBA.

Step 3: Import the Volume Groups and perform the VG activation.

Step 3: Perform VG Configuration Backup restore for those newly discovered VG/LVM, followed by VG activation.

Step 3: Restart the host to boot from newly mapped LUN that is part of LUN Import.

Step 4: Mount the LVM config on to the respective mount points.

Step 4: Create the new PCS resource groups and add the respective resources for the available LV and filesystems. NOTE: Please note that the newly activated LVM objects have the LV and filesystem structures intact.

Step 5: Verify the PCS Cluster status for all the resources are online and available status.