When deploying or thinking about deploying Cisco’s Unified Computing System (UCS), it is important to understand the concept of a service profile. A service profile is an idea that is central to the deployment of UCS as it is used to define the computing, networking and storage requirements of each deployed server. This article takes a brief look at the concept of a service profiles and how it can be used to reduce the amount of ongoing management of the UCS.
UCS Service Profiles
The first basic question to answer is: what is a service profile? The basic concept follows a virtualized server model. For example, when deploying virtualized servers in place of a physical server, a profile of the physical server must be created and then replicated within the virtualized environment. This profile includes the processing, storage, and network interface requirements for the server being deployed. The problem with this is that the management of the server and the network and storage (storage area network) management are all separate. The idea of the UCS is to wrap not only these profile items for the individual server but also include the ability to profile the required elements of the network and storage infrastructures and have those all provisioned at once within the same management system (UCS Manager Overview). The combination of all of these different items is referred to as a service profile.
The UCS service profile includes all of the following elements (among others):
- LAN Addressing
- SAN Addressing
- I/O Configuration
- Firmware versions
- Boot Order
- Network VLAN
- Physical port
- Quality of Service (QoS) policy
- Local Disk configuration
- BIOS versions and parameters
Obviously, the number of options that can be provided by defining various service profiles within the UCS offer the ability to easily provision new servers along with their network and storage requirements and connections. This is extended by the ability to monitor each of these elements within the same management console, thus ensuring any potential conflicts with elements like firmware are minimized as this management is monitored and maintained within the same system.
Service profiles extend their power through the use of templates. At its most basic, a service template can be used to create a new server with its appropriate resource requirements and with appropriately changed customization requirements (UID’s, server names…); this is referred to as an initial template. Another very useful ability of service profiles is their support for updating templates; for example, if a server is deployed using an updating template any changes that are made to the template (i.e. firmware) will be cascaded to all associated servers. As firmware management can easily be a nightmare for many administrators this takes the complexity out of the process.
There are three different methods to use when creating a service profile, these include:
- Manual – Involves the creation of a new service profile within the UCS Manager GUI
- From a template – A service policy can be created from a template
- By cloning – Involves the cloning of an existing service profile
The ability to create these profiles in so many different ways offers administrators even more flexibility within the system.
The concept behind the service profile is rather easy to understand by anyone with experience in virtualization and not particularly hard to grasp for those without it. The idea of grouping together multiple areas of the infrastructure under the same management domain is not a new idea but has only recently been offered under a single codebase. Hopefully this article gives enough of an overview to the concept; for those looking for additional information refer to the Cisco web page.