What ever happened to network management?
- 05 March, 2008 11:23
Enterprise Management Associates is back again after a year of absence, this time on a monthly basis. We've really missed your dialog and comments, and hopefully we've been missed as well. As before, this column will focus on network management trends as they fit into the bigger IT management landscape - focusing from time to time on application performance, or SLM versus BSM, or new approaches to asset management, or trends like CMDB adoption as they may affect the NOC.
But as an opening column, I'd like to target an area that has begun to haunt me, as well as some of our IT clients: What ever happened to network management?
This column idea was sparked by a query from an executive in operations working with EMA to try to figure out what a viable network management strategy should look like in 2008. This was not a naA¯ve individual, but one who couldn't sort out where to begin in planning to address network management in an environment so changed from what it had been even five years ago.
Let's just start by looking at the obvious marketplace phenomena. Five years ago, when many of us thought of network management, we looked at companies like SMARTS, Micromuse, Aprisma (Spectrum) and Concord as good places to begin. While there were many other significant companies out there, these brands were very much at the center of what defined the network management marketplace, along of course with HP.
But just in case you've been stranded in the Pacific for the last half decade, SMARTS was acquired by EMC late in 2004, CA acquired Aprisma and Concord (after Concord acquired Aprisma) in 2005, and Micromuse was acquired by IBM in 2005. Last year Quest acquired Magnum Technologies, a much smaller but still 'core" brand in network management. This leaves other small players in the event management area such as Entuity, CITTIO and Monolith, as well as more mid-market and SMB offerings such as Solar Winds and Ipswitch. Yet for the most part these vendors are overshadowed by enterprise brands in which the role of network management has become, to be blunt, problematic. In many of these platforms, politics have gotten in the way of effective integration, leaving network management and its potentially transcendent technologies (analytics, discovery, etc.) in a virtual limbo between foundation and stepchild.
There have been other bright spots, however, that may suggest a redefinition of what the "heartland" of network management may be. The most conspicuous is a focus on application management, in what EMA is now calling "Networked Application Management," managing application services over a network for QoE, diagnostics and optimized performance. This area includes technologies such as application trace, flows for application volumes, packet analysis, latency, transaction response time, QoE and RUM, and WAN optimization technologies.
One might argue that this already has become the new heartland for network management. Vendors in this market are far too numerous to honor fully here, but a few obvious names are NetScout (which acquired Network General), Fluke Networks, NetQoS, Network Instruments and in many respects OPNET (which acquired Network Physics technology). Compuware and Cisco play very well to this space as well.
Then there are the smaller innovators such as Apparent Networks, AdventNet, Coradiant, Network Instruments, Packet Design, Packet Trap, Valencia Systems, Wildpackets and Xangati. And there are a cluster of moneymakers such as Riverbed, Packeteer and Expand focusing on WAN optimization, and hard-to-classify but important vendors such as Shunra, Netcordia and Ipanema -- all of which could claim some fit to this market.
Another area worth noting is the rise of network change and configuration management - but this too is a story of many acquisitions - Opsware to HP, Emprisa to BMC and Voyence to EMC (all last year), with freestanding vendors such as AlterPoint and Intelliden, and a cluster of innovators and some newcomers such as Dorado, UpLogix (with its unique mix of out-of-band, config and monitoring), and ArcSight (with its focus on security).
Asset-oriented network management vendors are another area worth looking at, but so far the market here seems as fragmented as the rest. Rivermine, Avotus, Asentinel and Digital Fairway focus on telecommunications resource management. Planet Associates and Visionael focus on managing and planning for change, and in the case of Planet Associates there is a well-thought-out CMDB capability. While these are just a few examples, the leadership among network vendors in next-generation asset management (in which change and service management unite with financial planning) is still very early-stage.
Then there are the many new innovators in unified communications and VoIP, wireless, and integrated network and security management. The names and brands here are indeed far too numerous to mention. But if you look at the list of Interop participants, for instance, these tend to dominate on a volume basis.
So, what ever did happen to network management? In my opinion, from an overall market perspective, network management has lost its way. While much of the rest of the management industry has been seeking out innovative approaches to unifying technologies and integrating cross-domain technologies, network management as a market has done just the opposite -- it has fragmented into non-logical, opportunistic and sometimes tactical pieces.
It's sad to see that these this fragmentation is already undercutting the role of the NOC as the place where "things can come together" and interdependencies can be managed more holistically. All the sadder since this need to manage interdependencies across domains has become the defining set of requirements for the management industry as a whole.