dba

I shouldn't admit this publicly, but I have very little work to do these days. The reorganization of the DBA team within OIT has whittled my list of "DBA" responsibilities to installing Oracle software on servers and forwarding email requests for additional disk space. There's a reason I bring all of this up, though.

There are a lot of efforts under way at the University to cut costs and centralize services. In short, it is a mantra of doing more with less. That's admirable, but it has been so awkward in its implementation that I've been seeing a lot of dysfunctional side effects. People are working at the lowest levels to implement things as they are directed, but at the top it is not all coming together with the desired outcome.

So remember, I am someone working at the University without a lot to do. In recent weeks, I've been told that we have no money to invest in new software. I've been told that we can host databases for University departments with no limits on size, but also with no allocation of staff support, all without charge to the department. I've been told that we do not currently have any more disk space available to allocate to database servers. In short, we have no money, we have no more disk space, we have a surplus of staff time. We will not charge for the resource that is scarce; we will withhold the resource that we have in surplus.

At the lowest level, people are working toward the cause of reducing redundancy by centralizing services. They're saving money by not buying new software and hardware. But when you put it all together, you get a recipe for unbridled consumption and poor allocation of resources.

I received a request today from a web developer to create a new series of Oracle databases for a project--yes, that is the bee in my bonnet that started me on this tirade. In 2008, I did a study that determined such a setup would cost the department about $700 per month. Spending $700 of University money is as easy as sending an email to oitdba@umn.edu and asking for a new Oracle database, and we'll give you as much storage, RAM, and CPU as you'd like. (Ignore for now the obvious security risk associated with affording anyone the ability to make such a request.) A 2,000 user CRM app is the same as a 10 user sudoku game in our eyes.

I can't order software to help me in my job, but you can spend $700 a month for a database to manage your fantasy football team.

It's dysfunctional at the management level because it gives no consideration to return on investment for a project relative to other projects. It does not allocate resources strategically. It encourages unbridled consumption. It gives away for free the resources that are scarce and limits those that we have in excess. The individual charges that led us to making these low-level decisions are thwarted by the lack of an overarching vision in their implementations.

Posted by watt0012 at 1:39 PM | Permalink | Comments (0)

I've had a few other articles in the works for a few weeks, but I cannot seem to get my mind to the point of finishing them. Instead, the thing that is on my mind is the way that the University community is reacting to the changes that we must face. I've been following the auto industry's travails and I think the University is on an identical trajectory.

GM and Chrysler chose to ignore the reality that they were not competitive. What we are seeing with them now is the inevitable crash after years of running their businesses in unsustainable ways. Ford fared better because they realized they were headed for a brick wall a few years ago. They started correcting their course before the bottom dropped out of the industry. GM and Chrysler, meanwhile, refused to acknowledge that their ways of doing business were outdated and they've suffered for it considerably more than Ford.

At the University, at least in the little corner where I work, it seems like there is a similar refusal to accept that things have changed. I started in 2000 at the Carlson School of Management and moved to the Office of Information Technology in 2006. We did fine while things were good, but now that our funding sources are diminishing we are facing some challenges. I remember being amazed both when I started at CSOM and later at OIT at how lax the policy was on acquisition of hardware and software, and how careless people were with those resources. What is more amazing, or maybe it is more unfortunate, is the refusal to acknowledge and accept that we can't continue that way.

The way things were done at the University last year or last decade are no longer viable. We cannot throw money at problems any more. We need creative solutions to problems in OIT. Isn't that why we're in the IT industry? Adding disk space or buying a bigger server are not IT skills but they are common solutions to problems here. It seems like we reactively do these things because that is the way we have always done it. And tuition will go up and people will be laid off as a result.

It is a brick wall. The longer we avoid dealing with it, the harder it will hit. The game has changed, so the sooner we accept that the better we will be able to cope.

Posted by watt0012 at 9:53 AM | Permalink | Comments (0)

I've heard or read people using the word "planful" at the University lately. It's not really a word, but ignore that for now. One of the things I noticed in coming over to central OIT a couple years ago is the lack of planning in the allocation of resources and time. Is it too harsh of me to put it that way? It seems like everything is done as a reaction to something rather than as a proactive activity.

That brings me to my point. I would like to become more proactive in solving problems here. We react to problems when they happen. But I can say I think the challenges that we will be facing in the not so distant future are things like OIT acting more as a whole-University IT department rather than just one for the central administration, less money for new hardware, and data security threats that we cannot afford to ignore. A brainstorming session among a group of people would identify a lot more. So why can't we be proactive and start working toward solutions now?

The Google Apps/GMail project illustrates the point. Nothing against the people who are working on that project, but it is a reaction to stirrings by people in certain departments external to central OIT. I am sure it will be successful, but maybe it would have been better if we had predicted the need and started working on it before people started asking for it. It is not as hard as it might seem to predict trends and act on them. I would include that as a function of management strategy.

Google Apps could be considered a part of a mobile device strategy, since the apps are available from any device from a laptop to a netbook to a smart phone. We should have a mobile device strategy. We should have had one two years ago, when Blackberries, iPhones, and Android started showing up and Google was snapping up mobile technology companies. We can predict the trend of smaller, network connected devices from various activities in the industry.

On the Oracle database front, we know that demand for centrally-hosted databases is coming. We haven't begun to test Oracle 11g yet, but we know it will be in demand soon. We know that hardware budgets will be smaller, but we're not really planning for having less to work with yet. With as lax as some of our security practices are, I feel comfortable in saying that a breach is not only probable but imminent. We can be planful about these things so that we are prepared for them when they come. It will make our lives so much easier.

UPDATE (4-13-2009): I may have spoken too soon on the mobile strategy. It appears that there is one brewing (requires login). That is great to see and I look forward to following it. Why do other people get to have all the fun? :)

Posted by watt0012 at 3:23 PM | Permalink | Comments (0)