How to successfully reconfigure a Storage Foundation HA for Windows (SFW-HA) environment after converting from a physical server to a VMware virtual server.
|Article:TECH213992|||||Created: 2014-01-15|||||Updated: 2014-02-04|||||Article URL http://www.symantec.com/docs/TECH213992|
After converting a physical server to a VMware virtual server (P2V), the Veritas Cluster Server (VCS) services may not start and the storage may not be present.
The VCS services will not start because the MAC address may have changed after performing the conversion. The storage will also be unavailable until it is reconfigured for the new virtual server.
After successfully using the VMware virtual machine converter, several items will need to be reconfigured for SFW-HA to run properly. Below are items that need to be reconfigured.
It is possible that the Microsoft boot drivers might get re-enabled preventing SFW to start properly. If this occurs, please run a repair for SFW to restore the SFW boot drivers.
1. Run the Veritas Cluster Configuration Wizard to correct all heartbeat network communications for the new network adapters. As shown below in Images 1 - 8, refer to the Veritas Storage Foundation and High Availability Solutions Guide "Configuring the cluster" starting on page 105.
Here are some commands to verify the cluster is running properly:
Use the following command to display the state of a system across clusters:
- hasys -state
You can confirm the GAB membership by running the following command:
- gabconfig -a
- lltstat -nvv | more
2. After the cluster is running, all the services groups that have a NIC or IP resource will have to be manually updated with the new MAC addresses as shown below in Image 9 and Image 10.
The storage has to be reconfigured:
1. DMP or third party Multi-Pathing solutions must be uninstalled.
2. The VMDK methodology will also require a restore of the data to the new volumes created in the new disk groups.
3. The VMDG resources will need to be replaced with resource type of VMNSDG and VMwaredisks
Note: For Raw Device Mappings (RDM) methodology, the diskgroup will need to be converted to secondary. (See Image 11)
4. The MountV resource will need to be updated with the new VMNSDG resource name for the VMDg Resource name attribute. (See Image 12)
Note: Refer to VMware documentation for best practices and steps for how to create the new storage configuration for the virtual server.
General Best Practices:
1. Ensure testing failover of all service groups on all nodes
2. Install the latest Cumulative Patch for the version of SFW installed
3. Some suggested documentation for VMWare with converting from a physical server to a virtual server
Article URL http://www.symantec.com/docs/TECH213992