Tony Lock, originally published on The Register
There are some things in life that everyone just expects to function, almost without thought. You switch on the light, there will be electricity available; you turn the tap and water will flow; you pick up the phone, there will be a dial tone. Over the course of the last decade, much to the surprise of many who have long toiled in the industry, users now expect most, if not all, IT services to be similarly available whenever and wherever the need, or the whim, strikes.
Go back ten to fifteen years and things were different. Users were unsurprised when a system went down or could not be accessed. Happy no, and certainly they were quickly on the phone to demand service access resumption, even the problem was caused by one of their own actions. But people did not routinely expect any IT service to function without interruption around the clock, around the globe. And of course, if anything ever went unexpectedly amiss, the first reaction of every IT manager, without needing recourse to technology, was to blame “it on the network”.
Today, not only business users expect the IT systems they access, however infrequently, to produce the service, application or data at the drop of a hat. Much of the attention that has been dedicated to the IT infrastructure to help improve service levels has focused on the infrastructure, namely servers and storage, virtualisation capabilities and, occasionally at least, on management tools. But still, the impression I get when I talk to IT managers is that little focus is paid to the networking side of things. Outside of large organisations and specialist providers of managed services, it is fair to say that the knowledge of how to design networks for optimum performance and security is a skill set possessed by few.
Of perhaps even greater importance is that like IT in general network management tends to only become visible when something goes wrong, be that a service interruption or security breach. The fact that every organisation is at least as dependent on its networking infrastructure as on its servers and storage is not widely recognised by the application managers who think such things as the network should ‘just work’.
The technical and corresponding business reasons for good IT operational practices are, to a greater or lesser degree, now well understood by organisations of all sizes. But most of the attention tends to be paid to the patching and updating of applications, servers, desktops and laptops with storage and networking trailing far to the rear.
The critical dependency on the networking infrastructure and its active management is under-appreciated. Part of this may be due to the fact that networks are now much more highly available and have almost become taken for granted. But it is also the case that managing networks and the myriad of devices that comprise much of the network can be extremely complex to set up and manage.
As a consequence many small and mid-sized organisations often look to external suppliers to take on the burden of network infrastructure installation, configuration and sometimes even routine management. It therefore sometimes becomes what the author Douglas Adams termed an ‘SEP’, namely somebody else’s problem.
It is likely that this situation will now slowly change. As well as sound IT governance practice requiring that network devices be actively monitored and managed there is a growing raft of legislative and other external drivers making it clear that the status and configuration of such systems match pre-set standards or policies. It is, therefore, fortunate that the management tools available to help administer network devices are becoming more straightforward to use. Of course, that still leaves open the question of just what policies should be followed.
Another factor is the rapidly accelerating consolidation amongst the suppliers of network management devices and tools. To take just a few examples the last few months have seen HP acquire 3COM whilst the last few days saw IBM buy network device management specialist Intelliden. Over the same period network giants Cisco and Juniper have also been quietly adding to their capabilities and portfolio of tools.
With systems management tooling becoming available to help scan and administer network devices from a number of suppliers it will be interesting to see how actively the big vendors and their many channel partners try to educate organisations on the risks of poor network management and the potential benefits achievable through active device and network service governance. It could probably be argued that this is an area where all scales of organisations have room to improve.