Working Smarter, Not Harder

Profile Templates for Cisco Servers with UCS Manager from VMware

Many companies have VMware infrastructure on top of Cisco servers, and one of the most common consoles for managing Cisco servers is called UCS, or Unified Computing Systems Manager. UCS combines industry-standard servers, networking, and storage access into a single, unified system. This makes it possible to manage the infrastructure of all connected servers and control all of the parameters of the servers themselves, including tagging, configurations, BIOS, and more. The approach minimizes the number of connections between the server and the fiber interconnector and centralizes control of all of the Cisco equipment that is connected to it.

When a server is connected to Fabric Interconnect, one of the first steps is to implement management and basic definitions so that it will be ready for action. To do this, we need to build a server profile. UCS Manager makes it easy to capture or describe a server’s configuration details in a template that can then be used at-will to create vCenter Server profiles without needing to configure servers or define interface amounts from scratch repeatedly.

VMware offers a setup wizard for preparing a profile file that covers all of the policies, stages, sub-profiles, storage profiles, etc. It enables us to configure each component where desired. This file acts as a template and can then be copied to the target VCs to provide them with the same settings as the original VC. Further, after defining the configuration details of the original in the template, it is possible to edit and adjust the configuration settings that are passed on, if needed, without requiring the creation of an entirely new template.

When creating a profile for a server, it is preferable to know at the outset what the workload will be and for what actions the server will be used. For instance, when a server is intended for virtualization, some manipulations and customizations at the level of BIOS or CPU will help the server better cope with processes in the virtual environment. By specifying operating system best practices in the BIOS policy, the server will be able to optimize power management.

Likewise, we are able to configure memory management. With memory management in virtualization, we operate many servers and distribute many resources, some of which is memory. Intensive writing is then executed from the memory to the general servers in the environment since every virtual server gets its own allocation. Here, too, some level of manipulation of the workload is necessary to support a faster and more effective solution for writing. Instead of waiting for a “conversation” or communication between the operating system and the CPU or memory component to express that additional resources are needed (and then waiting for the hardware to confirm that resources are available), the server profiles make it possible to identify all of the cores – including virtual cores – automatically, making it easier to maximize performance.

The simplicity and flexibility of UCS make it an important tool for companies that employ multiple vCenter servers to manage Cisco servers. In addition to compliance challenges, it assists with the repetitive steps involved in provisioning and installing multi-VC setup, simplifying the tracking of VC configuration changes, and alerting to VC configuration drift. Rather than developing in-house automation solutions, UCS provides the answer enterprises need with minimal investment of time and skill.

Written by: Roman Kogan, Infrastructure and Virtualization Engineer