It’s good to see application performance across the network continuing to get mindshare in IT publications. Bruce Boardman, Network Computing’s management writer, recently took a look at a new approach to application management, Real-Time User Monitoring. I’ll refer to it as RTUM.
I think the RTUM concept is a good principle – to place the performance of IT applications and services in the context of their users. However, while some see RTUM as THE future of performance management, realistically it is promising for only certain niche situations and it will be adopted for those. It is not applicable for enterprise-wide network and application performance management.
My reasoning is that RTUM, as described in this piece, looks at one of the thousands of applications that run on top of the network, primarily Web traffic, i.e. HTTP/HTTPS. While many enterprise applications have Web front-ends for their applications now, this limited view will not be sufficient for network managers, engineers and operations staff that will need to look beyond Web traffic to diagnose performance problems of other critical applications. Consider the following cases that may not fall into the Web-based RTUM fishing net, so to speak:
- Complex or proprietary, home-grown applications often found in healthcare and retail enterprises
- Legacy applications like SNA or Banyan Vines common in government agency networks
- Latency intolerant, converged applications such as VoIP and associated call set up protocols increasingly found in many Fortune 500 corporations
- Highly volatile multicast trading applications that affect performance with microbursts in most financial firms’ networks
- Network blind spots caused by WAN optimization and security technologies, such as MPLS and VPNs the mask important application performance and user data
All that said, “context” will be key – while viewing RTUM results do you get the context of all the other applications and activity occurring simultaneously in the network? IT organizations engaging in troubleshooting exercises every day look at their tools to not just tell them user experience may be degraded, but rather forewarn with drill downs to pinpoint and resolve the problem before those users need to call a help desk.
The bottom line is that RTUM is not enough if you want to manage all applications flowing on your enterprise network.
Comments