Configuring PowerHA SystemMirror for cloud storage

Edit online

You can use the SMIT interface and the clmgr command to configure PowerHA® SystemMirror® Version 7.2.3, or later, to work with cloud storage options.

If you are not familiar with the processes for using IBM Cloud™ or Amazon Web Services (AWS), see the following websites:

Configuring Secured Shell (SSH) for SVC/DS8K storage

Configure the SSH protocol for each node that communicates with storage system. Public or private key pairs must be created on the cluster nodes and distributed to storage systems. PowerHA SystemMirror uses SSH protocol to manage the backup process on storage system. For more information about how to configure SSH for storage system, see the storage documentation.

Configuring the Boto interface

  1. To use IBM Cloud or AWS storage solutions, you must configure the Boto interface. For more information about the Boto interface, see the Boto 3 Documentation website.
  2. Configure authentication credentials in the Boto interface.
    1. Log in to the Identity and Access Management (IAM) Console
    2. If you have an installation of AWS CLI, run the aws configure command.
    3. Add the following information in the /.aws/credentials file for the corresponding cloud solution:
      Note: You must use the exact profile names that are specified in the following examples.
      Examples:
      IBM Cloud
      
[ibm_profile]
aws_access_key_id = YOUR_IBM_ACCESS_KEY
aws_secret_access_key = YOUR_IBM_SECRET_KEY
      AWS
      
[aws_profile]
aws_access_key_id = YOUR_ACCESS_KEY
aws_secret_access_key = YOUR_SECRET_KEY
  3. Configure the storage regions. You must update the storage region in the <user_home>/.aws/config file for the corresponding cloud solution to identify the cloud bucket is used.
    Note: You must use the exact profile names that are specified in the following examples.
    Examples:
    IBM Cloud
    If your bucket exists under the AP cross region use the following.
    
[profile ibm_profile]
ibm_profile_endpoint=https://s3.ap-geo.objectstorage.softlayer.net
    AWS
    If your bucket exist under the Asia Pacific (Mumbai) region use the following:
    
[profile aws_profile]
region=ap-south-1
Note: Ensure that your node time is synchronized with the cloud time to avoid out of synchronization errors. Use Network Time Protocol (NTP) to synchronize your node time with cloud time.

Restoring data from a backup profile

When you restore data from a backup profile, the target disk must be of the same size as the disk that was used when you created the backup profile. To know the size of your disk, run the following command:
bootinfo -s <hdisk>

The naming convention for the backup image file is: <cluster name>_<resource group name>_<volume group name>_<disk name>_<timestamp>.img.gz

For example: cluster11_RG1_VG1cbm_1G_2_hdisk5.2018-07-07_05:45:16.449614.mg.gz

To restore data from the backup profile, complete the following steps:
  1. Extract the compressed backup image file by running the following command. Do not perform this step if the backup image is not in the compressed format. 
    tar -xvf <image.gz>
  2. Transfer all data to the target disk by running the following command:
    dd if=/backup_file of=/dev/<target disk> bs=1024

    The backup data is now accessible on the target disk. However, the dd command might change the PVID of the volume group.

  3. Clear the PVID from the Volume Group Descriptor Area (VGDA) of the volume group and then reassign the PVID to the volume group on all cluster nodes by running the following commands:
    chdev -l hdisk$i -a pv=clear 
chdev -l hdisk$i -a pv=yes
  4. Import the disk into the existing volume group by running the following command:
    importvg -V <majornumber> -y <vgname> <hdisk>
  5. If the disks are duplicated from another set of disks that belong to a specific volume group, you might need to re-create the volume group by running the recreatevg command.
  6. If you used your backup profile to back up the rootvg volume group, you can verify whether the backup restoration was successful by changing the operating system bootlist to point to the restored disks and by rebooting the node.

Popular tasks

The following list contains popular task that can help you to configure and troubleshoot your PowerHA SystemMirror environment for cloud storage:
  • To run consistency checks in your cluster, run the clmgr verify cluster command.
  • To view the storage connectivity status for the backup profile, run the smitty hacmp command and select Problem Determination Tools > Storage Connectivity.
  • For information about the consistency group operations, image file creation, and cloud storage upload details, view the /var/hacmp/log/clutils.log file.
  • To view the status of the backup process and the schedule for when the next backup occurs, run the clmgr query backup_profile <backup_profile> command.
  • To view all files that were uploaded to the cloud storage service, run the clmgr query backup_files <backup_profile> command.
  • To verify that the connection to the cloud storage device is available, run the curl <address> command.