I’ve been asked a few times now for an example of what some network engineers are doing with our HP Network Automation ESX driver. What kinds of things might you want to do with the vSwitch, which is not a very complex switch? How could the example I used earlier, CDP being turned off, happen when it’s policy to have it turned on for all switches, and you use cool NCCM automation like you have with HP NA to check configurations for policy compliance before they’re deployed?

What if the server admin chose to have it turned off when ESX was being tested in the lab, which really wouldn’t matter, but then that vSwitch configuration was not changed for the production servers? Now the vSwitches have CDP turned off. That won’t show up as an issue because you can’t see the devices when you run your compliance checks. But every vSwitch is out of compliance, and can’t be discovered using CDP. Double trouble. With HP NA you can discover the devices, test active configurations against policy, and then autoremediate. I just spoke to a customer that routinely rolls out 20-40 servers temporarily at month’s end to handle peak application load for one of their applications. Those are tens of vSwitches coming into and out of production over the course of a week or so for that one application, every month. For that period of time, the network team has no visibility into them coming or going, or what state they’re in while they are in production. And if there’s a performance issue, those devices, which connect to the physical network via physical NICs, come and go. How would you like to debug that performance issue?

Let’s go one step further. Wouldn’t it be cool if the ESX servers were managed by a server automation (SA) product with heterogeneous platform support that integrated with the network automation (NA) product with broad device support, so that SA could notify NA when the VMs were deployed. NA would bring them under management, run policy checks, auto remediate issues, and update a shared compliance dashboard? That would be data center automation.  Shameless plug: hp.com/go/DCA. And if the NA product were integrated with fault, performance and IT process automation, then that would provide single-screen full network management automation. Another shameless plug: hp.com/go/ANLM. Why do I shamelessly plug our wares? Because the more I talk to network geeks, the more I hear that they had no idea how far we’ve come.