I just returned from Enterprise Connect 2011 (EC), which was previously known as VoiceCon. They changed the name because of the trend towards unified communications, video, and converged networking. I participated in two panels at EC.
The first panel was “Network Test Tools for Voice and Video,” a panel led by John Bartlett about the tools that are needed to support a voice and video deployment. In it, we were talking about how different tools have different strengths and no single tool does everything. An audience member asked how to identify the gaps in the tools that he has deployed, supposedly so that he can deploy more tools to fill those gaps. However, I’ve seen way too many sties that have an over-flowing tool box and they don’t use any of them. So my reply was “The tool that you have and use is infinitely more valuable that the five tools that you have and don’t use.”
My point is that you should select tools that you will use regularly for a specific function. Most sites start with a performance monitoring tool, which is conceptually easy. It should identify interfaces and devices that are running at high utilization levels or that are exhibiting errors (i.e. error counters are climbing). Configuration archiving and management are often next because incorrect configuration changes are the most common source of network problems.
The network management architecture that I use also includes event logging and network analysis. Event logging handles syslog and snmp traps, which tell me about real-time events that are occurring on the network. Network analysis tells me about network operational characteristics that are incorrect, such as a Hot Standby Routing Protocol (HSRP) configuration in which there is only a single router (there should be at least two to handle a failover event).
So that answers some of this audience member’s question about identifying the gaps in tool coverage. But I stand by my statement that any tool that you have and use is more valuable than any number of tools that you have and don’t use.
The second panel at EC was “QoS & Network Design for Converged Networks,” also led by John Bartlett. This was more like a workshop than a panel, because six of us teamed together to present an in-depth coverage of the topic. It was great to get audience participation through some good questions about the topics that each of us covered. I talked about network resiliency, which is critical to continuing operations of converged networks when (not if) failures occur. One point of the presentation is that with a good design, you can take half of the infrastructure down for maintenance while the other half is handling the production operations. This avoids the need for network outages to perform network maintenance and upgrade operations.
The exhibit hall was full this year. The economy definitely looks like it is coming back. There were a lot of interesting companies and products. It is a good place to take your specific operational or management problems to look for good solutions.
-Terry
_____________________________________________________________________________________________
Re-posted with Permission
NetCraftsmen would like to acknowledge Infoblox for their permission to re-post this article which originally appeared in the Applied Infrastructure blog under http://www.infoblox.com/en/communities/blogs.html
