Cisco UCS utilizes Service Profiles for server provisioning. These are logical profiles that are comprised of many smaller parts, such as BIOS settings, NIC settings, HBA settings, Firmware packages, boot policies and more. Creating consistency within your compute environment has proven valuable. Having the capability to add chassis and/or blades to my environment with just a few cables, and bringing these servers online with the required settings based on my profile is most valuable. Apply a Service Profile to a new, replace or relocated blade, and Cisco UCS takes care of the rest, provisioning as you expect.
Improvements to My Organization
Cisco UCS has reduced our physical footprint, drastically simplified management and created strong partnerships between engineering teams.
Room for Improvement
Software defects that result in false environmental alarms have been a pain point for us. These defects are not operational or performance impacting, but they do result in many hours troubleshooting to rule out any potential risks.
Use of Solution
I've been using it for years.
As long as everything is correctly designed and properly patched, deployment is a breeze with instantiation of VMs on-top of a configured UCS environment possible within just a few hours. This is aggressive scheduling, but it’s absolutely possible given the numerous options available for scripting and automation.
We hit a software defect once that caused a reload of some critical assets. This was immediately resolved and is the only true case of a stability issues I have seen.
It's been able to scale for our needs.
Customer Service and Technical Support
Cisco TAC is typically great to work with. UCS has a call home feature that will automatically open TAC cases on your behalf when issues arise. I’d recommend calling in critical cases to ensure timely response.
The initial setup as a first-timer can be overwhelming, but once you complete it, any subsequent setup is straightforward. The biggest thing is making sure you properly design the solution and develop a scalable schema. Take into consideration other environmental variables that require specific configuration, such as hypervisor BIOS settings versus bare-metal BIOS settings.
I recommend having someone experienced with UCS perform the initial design and deployment. This could be someone you have in-house, or someone you contract. You’ll want to make sure your schemas as setup properly, any unusual requirements are handled properly, and profiles are built according to best practices for your particular environment.
Cisco UCS changed the server blade game, converging network and compute into a single profile-based platform. Now with HCI, Cisco is converging storage into UCS as well. I recommend getting in contact with Cisco and one of their channel partners for a whiteboard session, design conversation and potential proof-of-concept. This worked well for me in the past. I have since been capable of designing and implementing Cisco UCS environments without aid from external resources, only asking for design validations.