In the modern IT environment there are a number of different constants that exists; one of these is that organizations are continually looking to reduce their costs. This can be done in a number of different ways, from reducing the number of new equipment purchases an organization makes to laying off people to “save” on support costs. Cisco’s Unified Computing System is an attempt to reduce these costs by integrating a number of different pieces of the data center network that would previously be separate both physical and managerially. The basic idea is to integrate the function of the data center servers, network and storage networks into a single system that is managed by a single system; this system is Cisco’s Unified Computing System (UCS) Manager. This article takes a brief look at this manager and what it provides.
The UCS Manager is a single system that is responsible for the provisioning and management of multiple systems, including the fabric interconnects (Switch), blade chassis, fabric extenders (connectivity to the fabric interconnects) and different blade servers. The UCS management software itself is embedded in the fabric interconnects; all of these different components mentioned above are treated as a single entity. Management is able to be completed through a GUI, CLI or through an XML API. Once the physical installation is complete a service profile is able to be created, the profile defines the different server, network and storage requirements of specific machines be them physical or virtual.
The idea is to have the server, network and storage administrators get together and help define the different profiles that exist or will exist within an organization. This service profile is then used as a template in the creation of the data center infrastructure from the servers to the storage network configuration. Provisioning will no longer require a separate process be run within the server, network and storage groups; it would simply require a single provisioning of a specific service profile. The UCS manager would then take care of reserving the appropriate resources which were laid out in the service profile.
The UCS manager is not limited to providing only the provisioning services for the system; it also provides a number of different duties including device discovery, inventory, configuration, diagnostics, monitoring, fault detection, auditing and statistics collection. It is also designed for high reliability, because the UCS management software is embedded in the fabric interconnects, the deployment of two interconnects provides the ability to lose one of these interconnects without affecting the connection to the UCS manager.
The idea of bringing together a number of different data center areas under a single management system is not a new concept. The problem of course was that integrating the different packages from different providers on different hardware platforms proved complex at best. The part of the idea that is different with this solution is that each of the different parts of the UCS infrastructure falls under the same codebase and is implemented on equipment that was built to provide this specific solution. It does this and also provides connection options to many popular SAN solution providers. Hopefully, the implementation of the solution turns out to be as easy as the marketing portrays; only time will tell.