There are two supported methods for migrating to different disk units (and one unsupported method is listed below as well):

Option A: Scratch/reload method
*Note: For migrating to different disk and a different frame/server, the only supported method is the scratch/reload method.

1. Perform two full system saves (The second save is in case the media is bad)

2. Power down the partition.

3. Remove the old external disk unit paths (For internal drives, you can physically remove the disk drives at this step. Otherwise, there may be additional performance considerations when running a mixed environment of both internal and external disk storage)

4. Map the new external disk unit paths
5. If your console is HMC managed, remember to update your partition profile so that its Tagged I/O load source is pointing to the new load source disk controller card and save the profile.
6. Scratch/reload from the full system save.

Option B : Add new units, move data from old units to new units, migrate the load source, and remove the old units.
 1. Add the new units to the System ASP (F10 to add and balance data), except leave one non-configured to become the new load source later.
 *Note: This can be performed with users on the system, but the extra disk activity will have a performance impact. The priority is static and set to *LOW to minimize the impact. If you plan to remove the old disk units immediately after the new disk units are added, then you can save time by selecting option 2. 'Add units to ASP' in step d. If you wait to remove the old disk units then performance issues will occur if the disks are not balanced.
https://www.ibm.com/docs/en/i/7.5?topic=pools-adding-disk-units-existing-asp
     a. STRSST and sign in
     b. Work with disk units
     c. Work with disk configuration
     d. Option 2. Add units to ASPs (without immediate balance) or option 4. Add units to ASPs and balance data. (*Note: do not add the unit that will be the new boot volume)
     e. Add units to existing ASPs
     f. Enter the ASP number next to your unit(s) to add back to the ASP, on the next screen, DO NOT press enter
     g. Press F10=Confirm Add and Balance data on units OR (Press enter to skip balancing)

*Note: If balance was selected, then wait for all of the new disk units to be balanced before continuing to the next step. If units were added without balancing accidentally or old units will not be removed from the ASP immediately, then you can issue the following command to balance the data to the new disk units until you are able to remove the old units from the ASP: STRASPBAL TYPE(*CAPACITY) ASP(1) TIMLMT(*NOMAX) PRIORITY(*MEDIUM)

 2. Remove the old disk units from the ASP besides the load source

 *Note: This can be performed with users on the system, but the extra disk activity will have a performance impact. The default priority is set o *LOW to minimize impact. If you would prefer to increase the priority (at the cost of a higher performance impact), then you can first STRASPBAL *ENDALC on the old disk units and then prompt STRASPBAL *MOVDTA with F4 and set your priority by leaving it at *MEDIUM or changing to *HIGH.

To remove the old disk units and have them balance on *LOW priority:
     a. STRSST and sign in
     b. Work with disk units
     c. Work with disk configuration
     d. Work with removing units from configuration
     e. Remove units from configuration
     f. 4 = Remove

 3. Perform a full system save

 4. Copy Load source data to new disk

     *Note1: Ensure PTF for Known Issue DT426716 is applied, which addresses load source copy failing if console is disconnected during the process.

     *Note2: For better performance, you can initialize and format the intended target load source disk at DST just before the load source copy (Initializing before the IPL to DST may only help with formatting but another initialize will need to be performed).  Work with disk units > Work with disk unit recovery > disk unit problem recovery procedures > Initialize and format disk unit

Steps to copy the Load source to a new disk through DST:
    a. B Manual IPL to DST
    b. Work with disk units
    c. Work with disk unit recovery
    d. Copy disk unit data
    e. On the 'Select Copy from Disk Unit' screen, select your Load Source disk unit to be copied (External Storage LUN id or PowerVS storage World Wide name can be viewed to match with the IBM i disk unit serial)
    f. Enter a '1' to select your non-configured unit that will be the new Load Source disk unit and press enter. If you receive a warning about PowerPC, press F10 to ignore and press enter to continue.
    g. Confirm Copy Disk Unit Data

    *Note: After the Copy is confirmed the disk will be prepared (initialized if needed) and copied. During the copy, system reference code C600450E indicates that the Copy Load Source is Progressing.  The prepare and copy only need to run for the actual disk usage amount, so it is common to complete well before 100% is reached. When the copy has completed, the partition will automatically IPL to DST on the new load source. The old load source will be marked (internally as XMD1 on page 00000000) and appear as a non-configured disk unit. At this point the old load source can be physically removed if internal, have its paths removed if external, or can be added to the ASP if needing to be utilized as any other non-load source disk unit.

 5. Power down the partition

 6. Update your Tagged I/O and boot volume setting
HMC:

If your console is HMC managed and this target load source disk unit resides on a different physical or virtual adapter, remember to update your partition profile so that its Tagged I/O Load Source is pointing to the new load source disk controller card and save the profile.

Tagged I/O: 

If this is an IBM PowerVS cloud Virtual Machine instance and you have more then 127 disk units, then open a PowerVS case and request the Tagged I/O Load Source be verified/changed to your new Load Source volume.

Virtual Server Instance Boot Volume: 

The PowerVS User Interface is not automatically updated by the load source copy through DST. Manually change your boot volume through the PowerVS Dashboard by selecting your virtual server instance, under 'Attached Volumes' select the name of your new load source to view the details where the 'World Wide Name' should match the serial number of your new load source, select the back button, click on the three dots to the right of your new load source, select 'Set as boot volume', on the 'Confirm boot volume change' screen click on 'Confirm'.

If this is a standalone LAN console managed system, then skip to the next step.

 7. Remove the old external disk unit mapping (For internal drives, you can physically remove the disk drives at this step. Otherwise, there may be additional performance considerations when running a mixed environment of both internal and external disk storage).

For External Storage:

Remove the zoning on the switch if using separate WWPNS, and/or remove the Storage Host Connections and luns if desired.

On your VM you will go to the storage and checkmark the correct old disk units and then select the three dots to the right of one and then select 'detatch'.

 8. Activate your partition on B Normal. HMC managed should activate by the profile (not the current configuration) so that it uses your new Tagged I/O Load Source setting.

 9. Remove Failed and non-reporting Disk Hardware Resources:
    a.  Start System Service Tools (SST) using the STRSST command.
    b.  Log on to System Service Tools using the appropriate SST user ID and password.
    c.  Select Option 1. Start a service tool
    d.  Select Option 7. Hardware service manager
    e.  Select Option 4.  Failed and non-reporting hardware resources
    f.  Select Option 4=Remove for any DMPxxx or DDxxx resources in the list (Unknown status with a ?).
    g.  Press F3=Exit three times and then press Enter to exit SST.

 10. Reset the disk paths non-disruptively:
     a. STRSST and sign in
     b. Option 1. Start a service tool
     c. Option 4. Display/Alter/Dump
     d. Option 1. Display/Alter storage
     e. Option 2. Licensed Internal Code (LIC) data
     f. Option 14. Advanced analysis
     g. Enter the following data and press Enter:

         1      MULTIPATHRESETTER

     h. Enter the following data and press Enter:
     Options . . . . .   -RESETMP -ALL

*Note:  The next screen appears to be a warning but it is a normal and should show ***RESET MULTIPATH UNIT PATHS TO NUMBER CURRENTLY ENLISTED***. As it mentions at the bottom to force the 'error', you need to press enter and issue the -CONFIRM on the options and press enter. You need to also include the - ALL as shown:
     i. Options . . . . .   -CONFIRM -ALL

Option C: Copy/Clone/Replicate through External Storage
*Note: Copying the load source and migration through external storage functions are supported on IBM external storage devices, but IBM i development has not fully tested on each device and therefore load source migration is preferred through IBM i disk copy functions as listed in Step 4 of Option B.

Although many clients have been successful using the external storage to copy/clone/replicate the units, this method has not been thoroughly tested by IBM i development. For this reason, this method of external storage migration previously was not supported for IBM i partitions, but now this method is approved as an acceptable option by development (starting as of 2024). However, if there are any questions or issues with this type of migration, then you will need to open a case with External Storage Support and not IBM i Support unless the questions/issues are with running the IBM i commands.

The general steps include: