Voice of the DBA

Category: Editorial

  • Secure Enclave Concerns

    We worry about security as data professionals. It’s a constant issue, though it seems that so many of the data breach issues come from simple mistakes. We have lots of automation, configuration as code, and more. These days we should have a protocol and automated process for setting up systems, file shares, and other ways of allowing others to access data.

    Awhile back we had attacks against hardware (Meltdown and Spectre), which required patches to systems to secure them. This was disruptive, but I haven’t seen any widespread security issues as a result of these flaws. Many systems were patched, though it’s entirely possible that there have been exploits.

    In a report this week, researchers have found a way to read data in the secure enclaves of some chipsets. That’s disconcerting because lots of secure software counts on the Intel Software Guard Extensions (SGX) protecting data. This is part of the Always Encrypted enhancements in SQL Server 2019, which I was happy to see added. These are supposed to help ensure you can protect data, even if you don’t have control of the physical machine, which is common in hosted and cloud scenarios.

    A new exploit could be a problem. Intel says it’s not, but they have an interest in saying that. The research looks complicated, and I could certainly see this is hard to do, but attacking a large environment, like a cloud subscription, might be something hackers try. Even if they aren’t successful, there may be performance issues with SGX calculations from a software patch.

    The point of this piece isn’t to alarm you. Likely this is more speculative than reality, but you should be aware there are potential issues. Encryption and security are complex topics, and I’m hoping Intel and Microsoft (and others) come up with better protections and mitigation for the issue. However, ultimately it’s up to us tech professionals to understand what possibilities there are and ensure we are taking every precaution we can.

    Steve Jones

  • Updating Certifications

    When I first started my career, certifications were hot. I worked on the Netware CNE and Microsoft MCSE certs. Nearly 30 years ago these certifications were controversial as plenty of awardees weren’t competent at their jobs in the real world. However, marketing worked and employers wanted these certs, so they mattered.

    Over the years I’ve worked on books for a few certifications and taken a number of tests. Partially as research and partially to test myself. I’ve personally found the certifications to be helpful because they focus my efforts in learning about a technology. I’ve found them to be maddening at times because the questions haven’t always felt very applicable to the world.

    That being said, since I write many of the daily QotD questions, and regularly get complaints on my work, I appreciate how hard it is to write good questions that relate to the real world.

    In any case, I never pursued the MCSA, MCDBA, or MCSD, and I’m somewhat glad I didn’t decide to start. Microsoft is retiring these broad, general certs as of June 30, 2020. If you are working on one of these, you have until then to pass exams. After that, you can’t earn one of these certs.

    Microsoft isn’t abandoning certs, but moving to role based certs. These should be more granular in nature, focusing on the things you might do for a job. Of course, some of these are retiring on Apr 2, so it’s a bit of a mess right now.

    The part that many people might not like about this is that these certs are all Azure based. I think that upsets some people, especially those that only deal with SQL Server on-premises. I think that’s fair, though I do think that most of the skills relate to both, after all, they do have a common code base. Azure doesn’t always mean PaaS, and the IaaS offerings with SQL Server are going to be what you get when you install the software on your own hardware.

    I’m not going to advocate for or against certifications. Like any training or educational achievement, a lot of the value is based on what effort you expended and how you present your work. If you want to earn an older cert, you need to hurry. If you’re looking for a new one that helps you stand out from others, you have some options. If you’re interested in pursuing a certification, let us know today.

    Steve Jones

    Listen to the podcast at Libsyn, Stitcher or iTunes.

  • Why Containers?

    I’ve been looking at and dabbling with containers for some time. I got more interested a few years ago as I saw the changes that Microsoft made to ensure container support for SQL Server, especially with the introduction of the WSL and Linux container support on Windows. This is one of those technologies that I think fundamentally changes the world, especially when we get good orchestration options, like Kubernetes.

    I was thinking about how the world changed as I read this piece with a little history of containers. It’s an interesting read, and eventually gets to the point of the title, the containers have changed the world. They have, and I think they will substantially change how we think about software in the next decade as more and more applications and systems move towards using containerization.

    Containers do change the idea of what we think of as a computer. It’s not quite a piece of hardware and some software. Now it’s just software and it can be running on all sorts of hardware, perhaps even at the same time with complex orchestration layered over a variety of hardware.

    What does that mean for us data professionals? I think this will alter the SQL Server world. This image of a Big Data Cluster is the beginning, with the same architecture likely appearing for “normal” OLTP workloads over time, though likely in some managed environment. I think hybrid, Azure Arc type of architecture, where we have SQL Server containers connected together and allowing us a highly scalable data platform.

    I suspect we’ll have lots of other software components, communicating across networks, each in its own little container. We’ll learn to build and update containers, and ensure communications between different components. I’m already seeing some uptake at Redgate in creative ways, and I suspect we’ll deliver more containerized solutions, as will many others.

    Containers are the future, more so than VMs, and I’d recommend you start learning something about them.

    Steve Jones

    Listen to the podcast at Libsyn, Stitcher or iTunes.

  • Data Modeling in a New World

    The world is changing in some fundamental ways as we recognize and deal with some of the social changes that are occurring regularly. We have had new countries appear and old ones disappear in my lifetime. Borders are redrawn and political systems change. Many of those are just data changes that we can import from some recognized source into our tables.

    There are other changes that might change the way we model our database schemas. One example is the concept of gender, which has changed from the simple male and female of my youth. I haven’t thought much of this, and many of you might not have as well. No matter what your opinion of gender fluidity is, you should still understand how this might impact software and your database systems.

    I was thinking about this while reading an article on designing forms for gender diversity. While the form is important, equally important is how we handle this in the database. A single character might not be enough, but even if you encode various values, what do you use? I haven’t seen a standard and likely this area of data capture is still evolving.

    I suspect that ISO/IEC 5218 needs to be updated, and until it does, I think many of us that work on schema models ought to think about being flexible and storing data in a list, perhaps in a lookup table of some sort. The UK has recommendations, but they don’t appear to have much granularity for the different ways people identify. New Zealand is slightly better, but this does appear to be a thorny question from the modeling perspective.

    Much of the reason we capture gender for is some marketing purpose. This allows data analysts to better understand behavior as a person’s gender sometimes provides some understanding of other behavior. While adding in new pronouns, titles, and genders can increase complexity, it’s also important from the standpoint of having data that is useful.

    With that in mind, it makes sense to not use hard coded constraints in tables, and ensure we have a flexible data model that handles the data we are likely to receive, including some sort of not applicable or prefer not to say answer. As data modelers, our goals should be to accurately capture and record data, and that means we need to adapt to a changing world.

    Steve Jones

    Listen to the podcast at Libsyn, Stitcher or iTunes.