Voice of the DBA

Category: Editorial

  • Republish: Making Guesses

    I’m skiing this morning in Crested Butte and then coming home, so you get a republish: Making Guesses.

  • Using Dynamic SQL

    There are many ways that you can structure your T-SQL code to solve complex problems. Throughout my career, I’ve learned different ways, and also some patterns to avoid. However, those change over time. For example, there was a time when heavily loading tempdb caused issues in the v6 era. As a result, I tried to avoid using them. These days, I’d consider them a good option to think about.

    One of the techniques that I see people trying often is Dynamic SQL. This is where you don’t write the code directly, but you write code that will put together a string of code to execute, and then you call this with EXECUTE() or some other mechanism to get your results.

    There are places where you might find Dynamic SQL useful. I saw an article from Erik Darling recently that gives you some situations where Dynamic SQL can help. These are often situations where performance is worse with a bunch of conditional code that is jumbled together. Separating out code into different procs or functions can help here, but that becomes a maintenance headache in many ways.

    Dynamic SQL isn’t something to just try without thinking carefully about the ramifications. SQL Injection is a real concern, though you can mitigate the risk with some careful coding. However, you need to be careful how you structure code, and don’t just concatenate strings together and execute them. Read up on good practices from others, and ensure you test your code thoroughly. In fact, this is one place where automated testing can become important, as you learn of different injection techniques, you can add new tests to ensure your code isn’t vulnerable.

    As you work with T-SQL, it’s important to continue your education, learn how the platform and language work, how performance changes between versions, and what options are to be adopted or dropped. Becoming a better T-SQL developer takes work and practice, and it’s an area in which we should all be constant students.

    Steve Jones

    Listen to the podcast at Libsyn, Stitcher, Spotify, or iTunes.

  • Viva

    I want holographic screens and meetings. Heck, I’d like Teams/Slack/Zoom/etc. to recognize multiple screens and let me separate out different cameras to different screens. I’d like to be able to see more than a tiny thumbnail of a presenter, or of other people that are in a meeting. That might be one of the things I miss most when I’m in these virtual meetings. I can often barely see others.

    Microsoft Viva is a new platform for remote work and collaboration. I’m not completely sure what it is, though there is a video with the cool, enticing holographic screens. It’s not an app, but appears to be some way to connect various information in an organization into Teams. It’s like an intranet from the 90s, but designed to help someone navigate their organization and work environment.

    I’m all for bringing together more data, and especially making it easier for people to do so. I love all the single-sign on we have at Redgate, but everyone produces so much data through tickets, feedback from customers, notes in chat apps, emails, announcements made everywhere, internally and externally, video updates, and more. I find it extremely hard to keep up, or to even know if I’m keeping up enough, too little, or too much.

    What I really need is some prioritization and triage for me. For me personally my role. That sounds like I want a personal digital assistant, which I don’t know I’d have confidence in. I suppose I wouldn’t have confidence in a human assistant until I’d trained them, so maybe this is a similar task I’d need to complete, whether it’s a computer or human. I’d have to build trust, and expectations about how to deal with different situations.

    With software, however, I’d worry about data privacy. I know this is still an issue with humans, but there is less scale for humans. They can’t learn and remember as much, and there is less secret access to about what my human assistant might know. After all, they know, or should know, if they tell someone something. However, what a digital system stores can be accessed by, well, who knows.

    At least I feel there is a potential problem here. Maybe I’m just worried, but many of the problems I see with computers have to do with the scale of data access. Both in the volumes of data someone can see, but also the variety of people that can see the data. Microsoft has already had some issues with trying to boost productivity for individuals. They’ve gotten complaints over what they store and report on, and more data might make that worse. Perhaps they are going to find a balance with insights and privacy, but I don’t want to be a test subject for this.

    Changing the way we work, helping individuals be more productive and accountable, while protecting their rights privacy is hard. I applaud Microsoft for trying something, mostly because I think remote work will continue to be a part of our lives and it’s not great. I suspect that this is like many of their new ideas, half-baked and incomplete. They want to release something, and it might not be much more than minimally viable. Whether they continue to invest and grow this into a useful tool remains to be seen, but I hope they do. I’d certainly like to have more useful ways to learn, collaborate, and communicate.

    Especially if they build holographic screens.

    Steve Jones

    Listen to the podcast at Libsyn, Stitcher, Spotify, or iTunes.

  • What’s a Lot of Money?

    The other day I was talking to a someone about the amount of money their business spent on a particular transaction. This person thought the amount was high, but I didn’t. We had an interesting discussion from there.

    Today, I’m curious what you think is a lot of money that your organization might have spent on some transaction. A piece of hardware, some software, maybe even another company. Are there times you’ve felt they overpaid for some item that you wouldn’t have bought?

    Most of the people I know work for someone else. Their concept of large or small values is often dictated by their own view of personal finances. While we might spent US$1-2k on a computer, many of us wouldn’t buy a $50,000 server for home.

    For friends that own a business, even a solo one, they often have a slightly different view of costs. A business might need to easily invest $10k in a project, a piece of equipment, or something else. They often find that spending money is less about the amount than the time it takes to recoup an investment.

    I wonder how many of you look at a $100,000 purchase and don’t see the value returned. Perhaps you’ve seen your company spend $10mm and you wonder if there was any rationale behind the purchase.

    Dealing with any amount of money is hard when you don’t necessarily have all the information, and you’re not used to spending at that scale. I know each time I’ve bought a bigger house, it’s been a bit of a shock and created anxiety. However, when I look back years later, I often feel differently.

    I wonder how many of you feel that way about purchases by your organizations over time.

    Steve Jones

    Listen to the podcast at Libsyn, Stitcher, Spotify, or iTunes.