Primary Storage

Originally posted by: Adam Young

One-way to ensure you have a safe environment to test your restore scenarios and your need for access to your data without interfering with your production virtual machines is to restore your #VMs from IBM Spectrum Protect Plus into a fenced network.

Network fencing, is a technology that isolates or “fences” virtual machine configurations, while allowing full network access. Fencing enables you to work with live instances of the same virtual machine configurations on the production network.

According to #VMware, typical reasons to use fenced networking include a need to clone one or more configurations numerous times, or when you have complex or difficult configurations and cloning is an easier route than repeating the setup. With fencing, engineers can run multiple, independent tests on a configuration deployed multiple times. Fencing is particularly useful when a developer needs to examine a bug without interrupting or stopping ongoing testing on a configuration and enables you to reproduce the bug later, if the developer is unavailable.

You can use fenced networking with restore jobs running in Test Mode and Production Mode, in IBM Spectrum Protect Plus. Running a VMware Restore job in Test Mode creates temporary virtual machines for development/testing, snapshot verification, and disaster recovery verification on an instantaneous or scheduled, repeatable basis without affecting production environments. Test machines are kept running as long as needed to complete testing and verification and are then cleaned up after testing and verification completes. Virtual machines created through Test mode are also given unique names and identifiers to avoid conflicts within your production environment.

Production Mode enables disaster recovery at the local site from primary storage or a remote disaster recovery site, replacing original machine images with recovered images. All configurations are carried over as part of the recovery, including names and identifiers, and all copy data jobs associated with the virtual machine continue to run.

This blog is a continuation of part three in a series, where we will focus on the following subjects:

Getting started with Spectrum Protect Plus (VMware and Hyper-V)

Backup and SLA's

Restore Scenarios (This Blog)

Universal Search and Restore

Spectrum Protect and Protect Plus Integration

Sizing a Spectrum Protect Plus Environment

How do you restore your virtual machines within a fenced network for data re-use with Spectrum Protect Plus?

You can use the following steps to perform this process, specifically within a VMware environment:

To create a fenced network through a VMware Restore job:

1. From the navigation menu, expand Hypervisor > VMware and click Restore.
2. In the Restore page, review the available recovery points of your VMware sources. Expand an entry on the Restore page to view individual recovery points by date.
3. Select recovery points and click to add the recovery point to the Restore List. Click to remove items from the Restore List.
4. Click Options. Set the job definition options:

   Destination

      Set the VMware destination.

   Alternate ESX Host or Cluster

   Restore Type

      Test or Production

5. Select the Alternate ESX Host or Cluster.

From the Production and Test fields, set virtual networks for production and test restore job runs. Destination network settings for production and test environments should be in separate locations to create a fenced network, which keeps virtual machines used for testing from interfering with virtual machines used for production. The networks associated with Test and Production are utilized when the restore job is run in the associated mode.

By default, the Use system defined subnets and IP addresses for VM guest OS on destination option is enabled. To use your predefined subnets and IP addresses, select Use original subnets and IP addresses for VM guest OS on destination

To create a new mapping configuration:

1. Select Add mappings for subnets and IP addresses for VM guest OS on destination and click Add Mapping.
2. Enter a subnet or IP address in the Source field.
3.  In the destination field, select DHCP to select automatically an IP and related configuration information, if DHCP is available on the selected client.
4. Select Static to enter a specific subnet or IP address, subnet mask, gateway, and DNS.

Note: that Subnet / IP Address, Subnet Mask, and Gateway are required fields.

6. In the Destination Datastore field, set the destination datastore for a restore to an alternate ESX host or cluster.
7. In the VM Folder Destination field, enter the VM folder path on the destination datastore.

Note: The directory is created if it does not exist. Use "/" as the root VM folder of the targeted datastore.

8. In the Script Settings fields, select an uploaded script and a script server where the script will run.

Scripts and script servers are configured through the System > Scripts pane. See Configure Scripts on page 140 in the Spectrum Protect Plus User’s Guide.

Pre-scripts and post-scripts are scripts that can be run before or after a job runs at the job-level. Batch and PowerShell scripts are supported.

Click Save to save the policy options.

9. To run the job now, click Restore. To schedule the job click Manage Job(s) and define a trigger for the job definition.
10. Once the job completes successfully, select one of the following options from the Actions menu on the Jobs Sessions or Active Clones sections on the Restore page:

Move to Production (vMotion) migrates the virtual machine through vMotion to the Datastore and the Virtual Network defined as the "Production" Network.

Clone (vMotion) migrates the virtual machine through vMotion to the Datastore and Virtual Network defined as the "Test" network.

For a video demonstration on this subject, click here.

You can try a fully functional 30-day Trial Software model at the IBM Marketplace and get detailed documentation for IBM Spectrum Protect Plus in the IBM Knowledge Center.