The Soul of the Server

A server goes down and you see your weekend go down with it. This fear runs deep in any engineer, and, when it happens, heads hit the keyboard.
No longer. 

(Photo by Mattias Diesel on Unsplash)

Being in charge of the entire workload of a company would have anyone’s head spinning. The servers that are under your command have to work like a well-oiled machine or the fury of 1,000 administrators will come crashing down on you. But you’re not the only one whose soul will be crushed.

Servers can feel your pain

Well, not exactly… but close. We have been operating in a way that is rapidly becoming outdated. NICs have MAC addresses, HBAs have WWPNs, and motherboards have UUIDs - if anything changes, welcome to a configuration black hole. Burned-in identities allow a data center to operate, but they don’t give the agility needed to make quick changes in an evolving situation. A production server goes down - what do you do? You grab another coffee and get to work. If those identities weren’t bound then you could pop, lock, and drop a new host right into the rack and migrate all your configs and licensing. That’s what we are talking about when we say the “soul of the server” – when the identity and policies that make those boxes blink are able to migrate from one host to another.

How to Save a Life

More like saving your professional and personal life, Cisco created a transient “soul” in their server hardware, or “body”. Thanks to this innovation, those darned BIOS settings and stubborn MAC addresses are able to migrate from one server to the next without being bound to any specific hardware. If your key server goes down, you will be able to migrate its “soul” to a backup “body” within minutes! Why have we not done this before? No one has defined a server through the eyes of the application - that is, until Cisco’s Unified Compute System –

Quit Being Stuck in a Box

And it’s not just about savings lives – thinking outside the box can create enormous opportunities.  Imagine upgrading server hardware by moving the soul from an old host to a new one.  Think about reducing downtime during scheduled maintenance windows by leveraging test/dev equipment, or leveraging the software nature to deploy new servers in mere minutes.  And what if we could tie firmware to the “soul”, thereby ensuring corporate compliance is always met even in the event of hardware replacement?  With UCS, all these things are possible.

What does that mean?

As software-defined infrastructures take the modern data center by storm, this original software-defined Cisco innovation is still leaving its mark in the lives of its users.  It’s about simplified management and recoverability in an age when zero downtime is the expectation.  It’s about abstracting policies that used to be bound to equipment so those policies can exist elsewhere.  It’s about freeing you, the administrator, to escape the confines of the data center so you can drive innovation.  Giving time back to you, the trench warriors, will not only save the company money but allow innovative business outcomes to thrive.

If you've reached the bottom you might also like CCIE Jeff Kish's, "The Golden Data Center" from VMUG UserCon in Indianapolis!