By Deepak Munjal, HP Networking Strategist
I am fortunate to have attended several of the recent Pathways to the Cloud roadshow events where I spoke about the network implications of moving to the cloud. I had a chance to visit cities across the U.S. during a very ice spring. Turns out New York and Chicago were a lot warmer than I thought and LA and Phoenix were not as warm as I was expecting. Having a late-night pizza in Times Square certainly made New York my favorite city of the roadshow. But most interesting were the discussions I had with the attendees.
Addressing issues on the pathway to the cloud
I had planned to spend most of the discussion on issues that network administrators would face when implementing server virtualization as they moved applications to the cloud. Many came to mind including:
- New “chaotic” traffic patterns forcing flatter network architectures
- Increasing traffic volumes requiring scalable L2 fabrics
- Network security in a virtualized world
- Multi-tenant requirements
- End-to-end automation
The white paper “Building Cloud-Optimized Data Center Networks” does a good job of addressing many of these issues and how to design networks that accommodate the migration to cloud. John Gray also does a great job going into more detail on the specific network issues that you’ll likely face in his blog “What is the impact of cloud on the network?”
What I also discovered from my experience in talking to attendees across six different cities was that while these issues were top of mind, there were also other topics that required discussion.
Reflections and observations: what was really on the minds of the roadshow attendees
I certainly enjoyed my conversations with attendees of the roadshow and especially the lively Q&A discussions after my presentation. With an audience mixed with server, storage and network professionals, I wasn’t sure what to expect. What I discovered was that they while they all understood how their roles would be affected by the move to the cloud, they all had one issue in common that they still had difficulty in addressing: provisioning new applications.
Because each group was involved in the provisioning of a new application, each was forced to use a serial process with management tools from various vendors to seamlessly deploy an application atop their individual physical infrastructure. The network portion was especially difficult because it required touching many devices from routers and switches to firewalls and load balancers.
Virtual Application Networks address new application provisioning challenges
HP addresses this problem with our announcement of Virtual Application Networks last week. You can read more about our Virtual Application Networks announcement. Mike Nielsen also addresses the topic in more detail in his blog “Reduce new application deployment time from months to minutes—with Virtual Application Networks.”
Virtual Application Networks are part of the larger HP FlexNetwork Architecture that is the foundation of all of our cloud offerings, including HP CloudSystem.
HP Virtual Application Networks is going to be a fundamentally new way of provisioning applications for the network and promises to speed the time to deployment from weeks and months to mere minutes.