IBM Support

Cloud Storage Solutions for i, BRMS full system save to the cloud

How To


Summary

This document will guide you through doing a full system save to the cloud using BRMS. The save will be done in 2 control groups, the first control group QCLDBSYSxx saves *IBM, *ALLUSR, DLOs, and IFS to virtual tape. The second control group QCLDBIPL01 does a SAVSYS and saves select IBM libraries and IFS objects to virtual optical. The virtual optical images can be used to boot from in the case of a full system recovery, from either DVD or for advanced users an NFS install. They also contain the IBM libraries and IFS objects that are required to retrieve the virtual tape images from the cloud for the remainder of the recovery.

Steps

Preparation:
 
This is a complex procedure. Prior experience with BRMS is important, and if you have not done a full system save with BRMS, performing a full system save to the cloud, and ultimately a recovery, may prove to be very difficult. Consider contacting IBM Lab Services for assistance.

First, be sure you have the latest BRMS PTFs and the latest Cloud Storage Solutions for i PTFs on your system:
Review the System Backup and Recovery from the Cloud page on the BRMS and Cloud Storage Solutions for i wiki before attempting your save. Keep in mind that future enhancements to Cloud Storage Solutions could allow it to work for larger system, currently it is only recommended for backing up 2TB or less, and note that the transfers may take considerable time to complete. From the BRMS Wiki:
While the IBM Cloud Storage Solutions for i using BRMS may be convenient, it may not be ideal for users or organizations that transfer large amounts of data from a backup. This solution is intended for small environments saving less than 2TB of data per day. Cloud connectivity was not developed as a replacement alternative for physical or Virtual Tape Library (VTL) devices.
Since you will be saving your entire system to virtual tape & optical, this will take up disk space on your system until the images are transferred to the cloud, so your system will need to be at less than 45% before you will be able to successfully save the system to the cloud.
Next, since we will be saving some items to virtual optical, which you may end up burning to DVD, we want to minimize the amount of data being saved to virtual optical. The following libraries can be large, particularly if you have journal receivers in them (either intentionally, or unintentionally):
QSYS
QGPL
QUSRSYS
QSYS2
QUSRBRM

Do a DSPLIB OUTPUT(*PRINT) on these libraries, QSYS should be less than 5 gb, others should be smaller than that. If they are larger than that, check for *JRNRCV objects, clean them up and consider using a different library for journal receivers. If any of these libraries take over 10 DVD size images (about 47gb), your save will fail because by default BRMS pre-creates 10 virtual optical images. The default of 10 can be changed (up to 75) if needed to accommodate bigger libraries, see Set Number of Turnkey Automatic Volumes to accommodate large libraries.
Before starting, you need to have already created and tested your resource for your cloud provider. If you have not, this document will show you how:
Using CPYTOCLD & CPYFRMCLD to copy data to and from the cloud
The backup consists of 2 control groups that must be run in this order:
  1. QCLDBSYS0x contains *IBM, *ALLUSR, *ALLDLO, & *LINK. This is the bulk of your save, and will be saved to virtual tape. Tape volumes are 30gb by default, but you can change this size to be up to 100gb.
  2. QCLDBIPL0x contains a SAVSYS and select IBM Libraries and IFS files, and will be saved to virtual optical. The goal of this control group is to save everything needed to get a full system recovery (scratch install) to the point you can retrieve virtual tape volumes saved in QCLDBSYS0x from the cloud.
It is critical that the control groups are run in this order because you need the BRMS information about the virtual tape volumes to be on the virtual optical volumes. This is different than a typical full system save where the SAVSYS is done first, then the IBM, user data, DLOs, & IFS are done next, but it is a requirement for a full system save to the cloud using BRMS.
The QCLDBSYS0x control group does not put the system in a restricted state, but for the purpose of this example, I am putting the system in a restricted state before starting. If you do not, you will likely end up with not everything saving as some objects will be in use.
The save:
If you have not already created and tested your resource using CPYTOCLD/CPYFRMCLD do so using this doc:
Cloud Storage solutions for i - Using CPYTOCLD & CPYFRMCLD to copy data to and from the cloud
If you have not run INZBRM *DATA to create the control groups, do so now.
First put the system in a restricted state, if you normally shut applications down gracefully before putting the system in a restricted state, do so now using your normal procedure.
Next end all subsystems: ENDSBS *ALL *IMMED
Use WRKSBS and wait until the controlling subsystem is the only one listed.
Next, start the first control group, this will save to virtual tape:
STRBKUBRM CTLGRP(QCLDBSYS01) SBMJOB(*NO) 
A successful backup will end with "Control group QCLDBSYS01 type *BKU processing is complete." If yours ends with an error, review the joblog to see what the problem is. Depending on the problem, you may want to resolve the errors and re-do the backup.
When you have successful backup, you can check the status of the control group using WRKSTSICC *ALL *ALL. Since your system is in a restricted state, your transfers will be in a failed status because the communication interfaces needed to transfer the files are not active:


image 11510

This is OK, we will transfer the volumes later. It is possible to transfer in a restricted state by starting bare minimum interfaces, but I chose not to because if you do, after each volume fills, it will transfer that volume, and your backup will be on hold until that transfer finishes, which will increase the total time required for your backup by multiples. I chose to let the transfers fail, then restart them after the entire backup is complete.

Next start the second control group, this will save to virtual optical:
STRBKUBRM CTLGRP(QCLDBIPL01) SBMJOB(*NO) OMITS(*IGNORE)

A successful completion will end with "Control group QCLDBIPL01 type *BKU processing is complete."

This backup may end with an error message at the bottom of the screen "Control group QCLDBIPL01 type *BKU completed with errors." This is likely because the QCLDIPL *LNK List (provided with BRMS) includes objects that may not exist on your system depending on what products and options you have installed and how your system is configured. It is OK if some of the objects do not exist on your system. These objects can be identified by doing a DSPLOGBRM, F18 to get to the bottom, see entries like this:

Date sent Time sent Message                                                  
6/09/21   22:22:30  Object not found.  Object is /QIBM/UserData/OS400/TCPIP/O
6/09/21   22:22:30  Object not found.  Object is /QIBM/UserData/OS400/TCPIP/R

Press F4 on each to get the full text. You can remove these items from the QCLDIPL *LNK List using WRKLBRM then option 2 on the QCLDIPL *LNK list, blank out the ones you don't need, press enter twice to save. Be careful not to remove the wrong ones. This change just needs to be made once, then future runs of this control group will not fail with these errors.

You can check the status of your transfers using WRKSTSICC *ALL *ALL OPERATION(*TOCLD).

image 11511

Depending on how quickly TCP started, the transfers from this control group may be transferring, or they may have failed if TCP did finish starting before the transfer was initiated, or some may be failed and some active. To restart any failed transfers (including the ones from QCLDBSYS0x, you can do STRMNTBRM MOVMED(*YES) or MOVMEDBRM.

You can also do:
WRKMEDBRM SLTCRTDATE(*CURRENT) SLTEXPDATE(*CURRENT) VOL(Q*)

To see the status of the volumes in BRMS. If the transfer to the cloud has completed, the status should be *ACT. If the transfer failed, or is currently in progress or if it failed, the status should be *TRF.

image 11512

Note, there may be other volumes in *EXP status, these were volumes that were prepared by BRMS in case they were needed. These volumes will be removed when maintenance is run.

Once all transfers are complete, you have finished your full system save to the cloud. To confirm your volumes made it to the cloud, use the web interface from your cloud provider.

Document Location

Worldwide

[{"Type":"MASTER","Line of Business":{"code":"LOB68","label":"Power HW"},"Business Unit":{"code":"BU070","label":"IBM Infrastructure"},"Product":{"code":"SWG60","label":"IBM i"},"Platform":[{"code":"PF012","label":"IBM i"}],"Version":"7.3.0"}]

Document Information

Modified date:
06 October 2025

UID

ibm16460459

Page Feedback