Issue 53 – Focus on Results
March 3, 2009
by Sean Hull
2009 is in full swing. Let’s put the struggles of 2008 and the last few months behind us, and take the downturn as an opportunity to dig in, work harder, and get creative in business.
In This Issue:
Here’s a mantra, it may be new to some, it may be old to others. Focus on results, not process. There are all sorts of ways we can apply this in technology, so let’s look at a few.
Buying new servers?
Are you looking at the brand, or the what, where and how? Does it perform as you need it to, does it fit within budget, can you get it delivered when you need it, are replacement parts easy to come by? What the failure rate of components? Brand may help you simplify this process, but it also may hide details, and drive up cost. Not always, but sometimes.
Buying new software licenses?
Of course you have to consider the DNA of your engineering team, but also the costs of those new licenses. Are all those fancy features being utilized? In my experience, a very large percentage, perhaps 80% or more of the obscure features of a database are not utilized by a particular application. What’s more many of the arcane optimization bells and whistles, knobs and dials are never turned. With that in mind, an open-source option may very well provide the same RESULTS for a much lower cost than the commercial alternative. Beware the team of salesmen in crisp black suits. Just like when your bank tries to sell you on a mutual fund or two, there must be money floating around to pay for and influence these people.
Buying consulting services?
Many have the knee-jerk reaction to ask for rates, right out of the starting gate. Without knowing anything about the project, and/or anything about this resource’s efficiency, the information is next to useless. First you must have a scope to know how long the clock will be running. And second you must know the efficiency with which this person solves problems. A better gauge would be to talk to past associates, references, or whomever referred the consultant, to find out if they got their money’s worth. In the abstract did the person (a) solve their business problem (b) do so within, below, or above the expected cost.
Another tactic would be to broach the subject of fixed fee for the project. Of course fixed fees involve discipline on your part, when hiring services. You have to have a good idea of the scope, or willing to ask hard questions about the problems that need to be solved, and get them answered up front. You also need to be able to stay focused and within boundaries, while the project is progressing, and expect that additional work and items outside that scope will cost some additional amount.
Time to Upgrade?
Wait a second, hold on there a minute. Not necessarily. Sometimes focusing on results even means *gasp* using existing, proven, and very stable versions of software. In enterprise software, those versions that have been out in the wild for many years aka older versions can often be much more reliable and stable. Remember enterprise software does not necessarily work like desktop software. If desktop software has a bug, it only affects one person, and you’ll likely find a workaround. With enterprise software, many many users are sharing those services, making bugs, outages, and upgrades much more expensive.
The push to upgrade, and get the latest features of course is the course of action the vendor will suggest, or even insist on. But that doesn’t make it the best for your business. Even developers get caught up in this, wanting to have the latest and greatest features available, partly so they can keep their skills up to speed, and partly to take them out for a test drive. But one thing your business does not need, and that is to test drive bleeding edge technologies. What you need is a solid, reliable, and results oriented technology infrastructure.
Focusing on results rather then process has its rewards. It doesn’t mean to ignore process, but on some level to not micromanage something into the ground. It may also mean focusing on what some technology does, and what it can deliver rather than getting caught up in how sexy and trendy this new technology is. The advantages are there too when weighing consulting services, but don’t forget to consider the whole solution that will be provided, when weighing hourly costs for those services.
Taleb’s book really resonated with me. One quote I liked: "people overestimate their knowledge and underestimate the probability of their being wrong". It permeates all human estimation of risk. The interesting thing that he illustrates in numerous examples in his book as that, knowing that we underestimate risk, we can plan for our own bad judgement. We can add checks that help us with our own fallibility. This as I see it, is his truest wisdom. He’s not saying that he has any necessarily less emotional and human relationship to risk, what he is saying is that our behavior in handling risk can be managed, and it is this way that we stand above the crowd. Brilliant stuff, and highly recommended reading.
I’ve been trolling through iTunes in the last few weeks, and have found some real gems. With that in mind I decided to add a new section to the newsletter to focus on interesting, and technology relevant shows that I find. By far the best one I’ve stumbled upon is Financial Times – Digital Business.
For instance the September 10th episode of this year talked about Mesh Collaboration, Globalization, and Social Networking. Definitely worth a listen.
You can read it online with updates almost everyday, and the print edition comes out on alternate wednesdays. The podcast you can find here.
I recently gave a live webcast for O’Reilly and Associates. The title was "MySQL Replication: Audit, Test & Verify". If you’re interested in viewing this webcast, the entire thing is now up on youtube for viewing. I’ve embedded the video into my blog here.
MySQL Replication is fairly simple to setup for the first time. However over time it can become troublesome. Errors can show up in the slave log files, or it can fail silently. We look at the caues, and also demonstrate how to identify differences, and fix them.
The Onion does it again, this time with "Apple Employee Fired for Thinking Different"
We’ve always been about open-source technology, integration & mixing commercial technologies such as Oracle with open-source ones such as MySQL and Linux. As open-source becomes mainstream, and more shops consider moving critical services to these technologies, we continue to provide assistance and expertise for these transitions. Whether it is performance testing and tuning, benchmarking, high availability or recovery, we can provide services for your specific needs.
Looking for a top-flight DBA? Visit us on the web at www.iheavy.com.