Cisco UCS service profile backup requirements and considerations

You can use your Unitrends appliance to back up and recover Cisco UCS service profiles and related configuration objects. In the event of a disaster, you can use this feature to quickly recover your service profiles, greatly reducing the recovery time objective (RTO) of reconfiguring your network and servers.

The Cisco UCS environment provides a “virtual chassis” that enables you to create and assign hardware profiles to individual logical servers. You can then bring up the logical server on dedicated hardware that you can easily migrate to another server in the case of hardware failure, or migrate between servers that do not require 24/7 up-time for efficient hardware reuse.

For UCS B-Series blade servers and C-Series rack-mount servers, allocation of UCS resources and hardware is managed at the domain level by the Cisco UCS manager. Each server in the UCS is a “logical server” that utilizes various resources as defined in the server’s service profile, and there is a one-to-one relationship between a service profile and a physical server. The service profile references hardware requirements, such as hardware identifiers, firmware, state, configuration, connectivity and behavior, but is completely separate from the physical UCS environment. Once a service profile is instantiated and associated with a given blade, rack-mount server, or server in a server pool, you configure a PXE server or map a bootable ISO image to the virtual-media CDROM drive to install the desired hypervisor or operating system (OS). See the Cisco document Cisco UCS Manager Configuration Common Practices and Quick Start Guide for details.

A service profile may be associated with a template and various policies. A service profile template can be used to quickly create additional service profiles. Policies can be used to enforce rules to help ensure consistency. For example, a boot policy defines how a server boots, including boot devices, methods, and boot order.

Because service profiles are essential to managing the servers in your Cisco UCS environment, it is important that you protect these configurations. Unitrends leverages native Cisco UCS data protection for profile backup and recovery, utilizing the Cisco XML API. Unitrends UCS profile backups capture all supported profiles, templates, pools, and policies in your UCS environment. For a description of each supported object that may be included in the UCS profile backup, see Identifying files in UCS service profile backups. Once you have a UCS profile backup, you can easily recover these items to quickly spin up your Cisco UCS environment in the event of a disaster, greatly reducing RTO.

The following requirements must be met to protect Cisco UCS service profiles, templates, pools, and policies:

●   The Unitrends appliance must be running version 9.0.0-13 or higher.

●   The UCS environment must utilize the Cisco UCS manager for resource and hardware allocation.

●   The Cisco UCSM firmware must be version 2.0 or higher.

●   The Cisco UCS manager must be turned on.

●   The Cisco UCS manager must be added to the Unitrends appliance with administrative trust credentials that support native backup and recovery of UCS service profiles.

●   Only application-level backups of your service profiles and other configuration objects are supported, since the UCS manager is not a server.

●   The Cisco UCS may be configured as a stand-alone system, or as a cluster to support failover in the event of an outage. Which IP and name you supply when adding the UCS manager to the Unitrends appliance varies depending on this configuration:

●   The stand-alone configuration consists of one physical UCS fabric interconnect that runs a single UCS manager. To add the UCS manager asset to the appliance, you must either supply the IP address of this node or, if DNS is setup in your environment, you can add the asset by node name only.

●   The cluster configuration is comprised of two physical Cisco UCS fabric interconnects, one active and one standby. A UCS manager runs on each. To add the UCS manager asset to the Unitrends appliance, you must either supply the cluster IP address or, if DNS is setup in your environment, you can add the asset by cluster node name only. Be sure to add the asset by cluster name or cluster IP. Do not use the IP or name of either fabric interconnect. With this approach, Unitrends can connect to the UCS manager regardless of which fabric interconnect is currently active.

Unitrends recommends running weekly or daily full backups of your UCS profiles, templates, pools, and policies. If your profile data changes frequently, you can schedule fulls to run throughout each day at any desired frequency. If you schedule the backup every few minutes, be aware that if the last backup is still running, the next backup is added to the queue and will be started once the last run completes.

After ensuring all requirements have been met, do the following to start protecting your service profiles:

Step 1:   Add the Cisco UCS manager to the Unitrends appliance as described in To add a UCS manager asset.

Step 2:   Run backup jobs as described in To create a UCS service profile backup job.