We just completed a very busy week on the exhibitor floor at Cisco Live, demonstrating Zenoss Cloud Service Assurance and talking with attendees. People flocked to our booth to get their picture taken with Zenny the Zebra, but they stayed to talk and share and learn about the needs of their own organizations.
I’ve distilled the conversations into five key lessons learned.
1) Care for Customers, not Boxes
Increasing virtualization makes it hard to keep track of customers. There isn’t an “accounting server” any longer, nor is there a “finance firewall.” Not one we can touch, anyhow. Everything is shared and it’s moving around. The traditional tool of enriching events with customer name just doesn’t work. What customer name would you put on a fault event for a VMware host running 20 virtual machine guests?
The impact? IT Operations can’t identify the customer based on the device name. And when operations doesn’t know who the customer is, IT gets categorized as an unresponsive, uncaring cost center. And the business thinks about cheaper alternatives. Not good for IT, and definitely not good for the business.
The people I talked to feel this pain intimately. And they immediately identified the value that Zenoss provides with services. Add the customer’s virtual machines to a service and Zenoss automatically tracks the entire supporting infrastructure specific to that customer. If there’s a fault, we know how to fix this specific customer. Fix the customer first, the infrastructure second. That’s the way to win friends!
2) Integration, Integration, Integration
I was very surprised at the number of detailed questions about feeding fault data into other IT management systems. Of course people care about ticketing integration, but they need so much more. It’s critical to pick specific, meaningful events to integrate. They want to separate events by technology stack, customer application, and type of fault.
Before Cisco Live, I’d never demonstrated the Zenoss trigger and notification system. When we introduced it in 2011, I knew it was a major advance, very flexible and really simple to use. And after going through a dozen different scenarios on the show floor with as many people, I know how important integration is for IT success!
3) The Rise of the Infrastructure Engineer
Over the past 20 years, every time IT has added new technology we’ve added a new kind of administrator. Check around – you probably know a Windows Server administrator, a LAN administrator, a Linux administrator, a storage administrator, a virtualization administrator. A title for every kind of equipment – and a new set of tools to buy, too!
For the first time, I’m seeing that process reversed. When organizations install converged infrastructure (whether they’re actively building a cloud right now or not) they’re simplifying the engineering roles. One person, an Infrastructure Engineer, is responsible for the compute, virtualization, storage, and network stacks of the converged infrastructure. The engineer has broad responsibilities for driving customer satisfaction in addition to technical excellence. And the idea of jumping between four or more technology stack-specific point tools to get their job done is a non-starter. At least a third of show attendees already had people in the infrastructure engineer role and another third were very familiar with the concept.
4) Sympathy for the Event Flood Devil
I walked all around the Cisco show floor and no monitoring company was showing off a console full of red events. But every person who looked at my bright red demo screen nodded in sad recognition. It’s an all too familiar sight, and no one thought their organization did a good job managing issues to resolution starting from an event.
We demonstrated how Zenoss handles event floods with automatic cross-technology stack root cause analysis, infrastructure alert consolidation to customer impact notification, adjusting root cause analysis in real-time as corrective actions are executed. People were fascinated, but skeptical. “How do I set this up? What do you discover, and what do I have to build?” We showed them, adding new services and seeing how Zenoss built and maintained the cross-stack linkages dynamically, and learning how our policy definition was resistant to the entropy that destroys traditional rules-based systems. Were they convinced? Well, not until they run it in their own data center, I’m sure. But, I really liked the number of people who took the mouse from me and did it themselves!
5) Cisco UCS the King of Converged Infrastructure
It’s been hard to miss the message that Cisco UCS is gaining ground quickly. #3 blade server vendor worldwide. #2 in North America. More than 10,000 customers in just two years. Talking to people at the show, I recognized how real this is and how big the momentum is. Nearly every person I talked to was running Cisco UCS in their shop, replacing IBM and HP blades. Sure, we’re at Cisco Live and you’d expect it to some degree – but the progress, the mindshare, the attention was overwhelming. Looks like Zenoss was right to get in early (first!) with UCS support two years ago.
Bonus Lesson! Zebra Impersonator Etiquette
Our superstar zebra impersonator, Travis Balinas, spent Monday and Tuesday dressed up as Zenny the Zenoss Zebra. Hundreds of people lined up to get their pictures with Zenny added to our Facebook album.
I asked Travis what he learned in his two days of sweating, and he shared this key advice that I’m sure you’ll find useful every day. “If you’re dressed up in a zebra suit, don’t talk to people. It confuses them. Zebras aren’t supposed to talk.”