Network

Edit online

The network in your environment must already be configured for the existing resources that include hosts, HMCs, VIOS, and storage devices. The VM Recovery Manager DR solution requires that the KSYS node be directly connected to the HMCs and the storage controllers in both the sites. The KSYS uses the HMC to interact with the hosts and VIOS, and the KSYS uses the storage controller to interact with the storage subsystem.

You can modify the KSYS system properties to enable or disable the network mapping function globally across the sites. By using the network mapping function, you can create VLAN ID or virtual switch mapping policies that contain a mapping of VLAN IDs or virtual switches that are assigned to the virtual machines when the virtual machines are moved from the active site to the backup site.

The recovery operation is not easy with the default settings that are used in the AIX® operating system because of following issues:

Table 1. AIX default setting issues for recovery of the virtual machines from the active site to the backup site
Issue Resolution Related information
Data, which includes the AIX root volume group, is replicated from the active site to the backup site. The AIX root volume group contains information about the disk, adapters, network, and other system-specific items. The data also includes the storage area network (SAN) World Wide Names (WWNs) that identify each SAN disk. Although the target disks are exact copies of the source disks, the disks are different. Therefore, the AIX operating system creates new devices when it starts. For example, if the source virtual machine has 10 devices, hdisk0 - hdisk9, when the virtual machine is moved to the backup site, the hdisk0 - hdisk9 devices are marked as defined and the hdisk10 - hdisk19 storage devices are marked as available. In this case, the original disks, hdisk0 - hdisk9, are called ghost disks. These renamed disks might cause issue, especially when some scripts or tools depend on the original disk names. On the source LPAR, set the AIX ghostdev parameter to yes to clean up the ghost devices. For more information about setting this parameter, see Configuring VM Recovery Manager DR.
The naming convention of the disks that are used by the recovered virtual machine (hdisk0, hdisk1, and so on) might not match with the names of disks of the source virtual machine for various reasons. For example, the hdisk5 disk in the source site might be displayed as hdisk7 on the target site. This mismatch can cause difficulties in a disaster situation because the disks that were part of that volume group must be known to import any of the AIX volume groups. On the recovered virtual machine in the backup site, run disaster recovery scripts that address the following issues:
  • Set the local host name
  • Set the network IP addresses
  • Customize the host bus adapters (HBAs)
  • Import the volume groups
 For more information about the disaster recovery scripts, see Running the disaster recovery scripts.