The culture gap in your IT room

Last time, I shared with you a letter from a correspondent in South Africa who is required to restructure what seems to be a large and fairly complex IT environment.  His shop includes mainframes, open systems servers, and Wintel servers that are distributed over several campuses.   As he began to create his new storage strategy and consider his future hiring needs, it became clear to him that some fundamental differences exist between the way his Unix, mainframe and Wintel staffs went about their business.  Even though the teams had a similar focus and did essentially the same thing – their first responsibilities are for backups and recoveries – he began to appreciate that the differences between the groups were profound.

Last time, I shared with you a letter from a correspondent in South Africa who is required to restructure what seems to be a large and fairly complex IT environment.  His shop includes mainframes, open systems servers, and Wintel servers that are distributed over several campuses. 

As he began to create his new storage strategy and consider his future hiring needs, it became clear to him that some fundamental differences exist between the way his Unix, mainframe and Wintel staffs went about their business.  Even though the teams had a similar focus and did essentially the same thing – their first responsibilities are for backups and recoveries – he began to appreciate that the differences between the groups were profound.

No one has, to my knowledge, done a study that looks at how the cognitive processes differ between IT-ers in these three groups.  Nonetheless, I am pretty sure that the way mainframe, open systems and Windows staffers approach their work has some marked differences. 

Part of the reason may lie in the fact that many mainframers and open systems managers go back to the days when command-line interfaces were all that existed, and they have preferred the familiar “glass TTY” metaphor ever since.  They often distain the snazzy GUIs that Windows managers have come to expect because “it puts them one step further away from the systems they are trying to manage.”

Most guys who manage Windows machines grew up with wizards, and see no reason to revert to the primitive IT processes of their brethren who work on the larger systems.  They point to the fact that with many wizard-based systems, any booby can learn how to manage a system in a hurry.  Mainframe people often agree.

End-users of the various machine types typically have different sets of needs as well, but that has to do with the different applications that reside on the machines rather than with the machines themselves.

Most IT needs are fairly common across all systems and although the implementation of how those needs are addressed is often quite different, it clearly makes good sense to have a single product capable of performing some needed action across all platforms. For example, it is certainly smarter to have a single backup and recovery product that manages all your data wherever it resides than it is to spend money on separate solutions for mainframes, for Wintel, and for your various flavors of Unix.

But if the requirements of the managers and users are different then it stands to reason that the most efficient enterprise management product must address the differing needs of all groups at once. 

In a circuitous way, this actually does come back to my South African friend’s letter.  He envisions bringing his several teams, currently working in isolation, into a “single storage management team who have a global view of all storage that exists.”  

Probably the advice we should offer him is to maintain his existing teams (and the knowledge bases they represent) intact, but to install a cross-platform group tasked with aiding knowledge transfer between the groups and with building some enterprise-wide policies that address corporate issues in a rational way. 

It is, however, much easier to philosophize about this than to actually do it.

Meanwhile, if any psychology grad students read this newsletter, there is a darn good dissertation topic out there for anyone who wants it. 

And as for the rest of you, “Lekker naweek!!” which my friend in Johannesburg assures me is Afrikaans for “happy weekend.”  Of course if he is one of those Unix guys, who knows if he’s pulling my leg?