Please elaborate on that last point you just made about the need to have virtual machines talk directly to networks…
Ill give you a parallel in automobile support and servicing to show you what were talking about.
I have a car and interestingly it knows more about how it drives than I do. So when theres a problem. I get a warning light that pops up on the dashboard which basically says: Go and get me serviced. I have to then take it in for servicing. The technician tells me whats wrong, and then he fixes it. If you look at this chain of actions, you can see that if theres a problem with the car, the biggest amount of time from when the warning light comes on until its fixed, is the time it takes for you to drive it to the service centre and get someone to look at it.
Now were talking about how we can see it being the same when it comes to service and support for your network. When theres a problem with your network, its one thing to have a red light come on, or something like that, but the time it takes for you to extract all of the information, send it to your vendor, wait for them to process it before coming back to you saying, OK, its the problem is…, and then proceeding with the fixesthat sort of time actually can translate into significant outages in your network.
If it were your car, you wouldnt be able to drive it.
So ultimately what weve presented here with these announcements is not a proposition for basic efficiencies when it comes to driving out costs in the capital expenditure. Were also talking operational costs, and improving things like the time to repair a network. After all, if there is an issue, why dont we have the network just call the vendor directly, tell it its got a problem? More often than not, we know the issues before our customers have had a chance to work it out. Weve done a lot of work in customer care particularly in the areas of automation focused on reducing the amount of time it takes to repair something, or perform tasks where automation can really help, such as those in the management of virtual machines
Apparently, when our competitors talk about a technology called VBlock, which is about the management of storage and computing resources, one thing they leave out is the question, How do we manage things like network resource? We dont want to leave that out. One of the announcements we have made is about how, through a partnership with VMWare, Juniper has produced an orchestration platform that interacts directly with and control virtual machines within the data centre.
Sign up for Computerworld eNewsletters.