ra2 studio - Fotolia
Public cloud management tools lacking, research finds
Networking experts discuss the challenge of finding robust public cloud management tools, how to improve network resiliency and avoiding network exceptions.
Thinking about moving applications to the public cloud? If so, you better also think about the management and monitoring tools you'll need to keep track of those assets. The bad news? Finding the right public cloud management approach isn't a simple task.
Shamus McGillicuddy, an analyst at Enterprise Management Associates in Boulder, Colo., recently detailed EMA research that found 72% of 250 network managers said they need new public cloud management tools to oversee their public cloud deployments. Sixteen percent said they were using incumbent tools, while 14% said they were still looking for ways to satisfy their cloud management requirements.
"This is not a trivial adjustment," McGillicuddy said. "Tools are often the foundation of the network team's established processes and best practices. When the cloud forces them to acquire new management solutions, there will be pain."
Several issues contribute to the public cloud management challenge. First is securing access to public cloud services; legacy hub-and-spoke WANs based on a central security foundation don't perform well in a cloud-centric environment. Monitoring and troubleshooting internet-based cloud connectivity is another problem, McGillicuddy said, citing EMA's research.
Read the rest of McGillicuddy's discussion on public cloud management challenges.
Making the network resilient
Network engineer Brian Keys took a look at network resiliency and why it's so difficult for enterprises to have a network that's highly available. For one thing, nobody wants to pay for the technology necessary to achieve that goal. Additionally, finding architects with the experience to design a highly available network isn't easy.
Still, Keys said, enterprises can take steps to improve their network's reliability. The use of uninterruptable power supplies is a good approach. So are redundant links for branch office connectivity. But knowing which techniques are necessary and which ones are just nice to have requires careful study.
"A competent network designer should be able to tell with a high degree of certainty just how resilient the network is and in which ways," Keys said. "Probably the toughest part is to explain to upper management the pros and cons of the new proposal and get their buy-in."
Read what else Keys has to say about network resiliency.
When to accept an exception (Hint: Hardly ever)
Ivan Pepelnjak at ipSpace had some thoughts about the wisdom of making exceptions in your network's design or operation. The conclusion: Don't. Or try not to.
Recounting an experience he had as a passenger of an airline that had decided to delay his flight while waiting for a late transfer, Pepelnjak said while the intent might have been good, the result wasn't. His flight -- and many others -- were delayed for hours because of airport congestion. If the airline hadn't waited for the incoming flight to land, his trip would have left on time.
Network operations face some of the same challenges, he said.
"How many times did you modify your network design or implement a one-off solution to accommodate an exception request coming in at the last minute? How many times did you mess up your network trying to do that?" he wrote.
While it's necessary to be flexible, network operators also must weigh the risk involved with accepting an exception. Make sure you remember the wider effect before considering a last-minute request.
Read Pepelnjak's thoughts on network exceptions and how good intentions can go wrong.
