Voice of the DBA

Category: Editorial

  • Stress Testing

    Many of the DBAs that manage production systems will at some point determine what level of hardware is needed to support a workload. Whether this is a physical server purchase or a cloud “rental”, someone has to decide what hardware is needed. How many cores, the amount of RAM, the number of disks, which hopefully correspond to some level of IOPs, and more. Even in the Azure SQL Database world, you must decide what database capacity you will pay for.

    Since this is a big decision, and changes can be hard to make, many DBAs overbuy hardware. After all, no one wants to have a slow server. This is true for Azure as well, at least for many people I know. While changing from an S3 to a P2 is quick and easy in the Azure portal, it’s not such an easy sell to management. If they’ve budgeted $150/month and you tell them we want to go to $900/month, the technical change is the easiest part of this.

    As a result, I’m surprised that we don’t really have better ways to determine if hardware will support our workload. I see this question asked all the time, and although there are tools and techniques suggested, I’ve yet to see many people have a set, known standard way of evaluating hardware and a particular workload.

    One one hand, I think there should be better tools to do this, whether from Microsoft or someone else. I suspect since this is such a rare activity and businesses have been willing to overbuy hardware (or deal with substandard performance), that there isn’t any large impetus to solve this issue.

    However I wanted to ask if any of you actually stress test hardware? Either your current hardware or new purchases. If you don’t know what your level your current hardware performs at, how do you compare that to new hardware?

    Do you have a way to replay and measure a workload? Do you have the time to do so when new hardware arrives? Is there a documented method you use? Apart from discussing this today, I’d love to see some articles that detail exactly how you test hardware from a technical tool perspective, and then a followup that examines and evaluates the results.

    Steve Jones

    The Voice of the DBA Podcast

    Listen to the MP3 Audio ( 3.2MB) podcast or subscribe to the feed at iTunes and LibSyn.

  • Losing All Traces of Data

    I was reading a thriller recently, in which a businessperson had their child threatened if they didn’t get some data for the criminals. Once the person had retrieved the data, they were told to delete it from the system and from all backups. Of course, they could do this, all in a few paragraphs of a novel. I’m sure plenty of lay people read this passage and accepted it as a possibility

    While I certainly understand how a user might be able to delete data from a system, especially in many third party applications that are poorly written but have sold well. However, could someone actually delete all information from backups? I’d say that in most of the companies I’ve worked in, this wouldn’t be possible. If the information was of any age, it would be stored in multiple locations on different media, some of which would be offline.

    However I haven’t worked lately in some enterprises where companies have moved to using disk backups, with systems connected together and managing versions. I suspect that it is possible in some of these enterprises to actually remove all traces of data from the organization, which isn’t what I’d ever want possible. If for no other reason than this is an incredible attack vector for ransomware, a malicious virus, or some other destructive process (including rm -rf). There’s also the issue of mistakes made by users; should they be able to remove all traces of data?

    There may be valid reasons to remove all copies of data from an organization, especially when some time has passed. However I think this should be a difficult process, have some hurdles to overcome, not the least of which is physical access, and should require multiple people to approve the actions. As we connect more and more systems, and rely on data being available, allowing anyone to permanently remove data without oversight will become a problem.

    Steve Jones

    The Voice of the DBA Podcast

    Listen to the MP3 Audio ( 2.8MB) podcast or subscribe to the feed at iTunes and LibSyn.

  • Cloud First

    SQL Server 2016 is the first “cloud first” release of SQL Server, as told to us by Remus Rusanu, former Microsoft SQL Server development team member. The features and enhancements in SQL Server 2016 have been running in the Azure SQL Database cloud for some time before they will be packaged up and released as the on-premise SQL Server product that most of us expect to work with.

    There are a number of items about this worth discussing, but one stands out to me. The idea of using large numbers of feature flags and deploying code to the “cloud” for use by customers and internal Microsoft people is interesting. On one hand, it’s really just a faster way of having beta software examined by users other than developers, with no install/uninstall/upgrade for users. Speed is good, but being on the bleeding edge and having systems break isn’t necessarily what any of us want. However the use of these flags to turn features off quickly means that disruptions can be minimized for individual customers.

    Some of the engineering process changes made to be cloud first were important for Microsoft to have one main branch of code for SQL Server. Having a separate branch for the cloud and on-premise versions had to be inefficient and problematic. However, that’s not something we, as customers, care about. We just want the database platform to work, wherever we run may run it.

    I do think that having features come out in Azure, whether private testing, public preview, or general availability is good. The people that can test these features give feedback quickly, and the rest of us aren’t affected by the problem code. More importantly, the developers at Microsoft get the chance to learn more about how the features will be used and tune them before a very wide set of on-premise customers get code. Personally I was glad to see things like Row Level Security (RLS) appear in Azure first (Mar 2015), and come later to the on-premise version (SQL Server 2016).

    I really expect in the future that almost every feature that is added to the SQL Server platform will be run in some form in Azure before it’s released for on-premise use. In fact, I’d be surprised if features are added to the on-premise without 6-12 months of testing in the cloud. That’s good, and I hope it leads to an increase in code quality for those of us that will run SQL Server on our own hardware.

    Steve Jones

    The Voice of the DBA Podcast

    Listen to the MP3 Audio ( 3.0MB) podcast or subscribe to the feed at iTunes and LibSyn.

  • Specialist or Generalist

    What would you rather be: a expert specialist in some narrow area of SQL Server or a generalist that’s competent in many areas? The same question could apply to any area of technology, like C# or Windows, or even any subject in which you work. What’s your preference?

    Most of us don’t have to make a strict choice. Often in pursuit of deep knowledge in one area, say indexing, the effort will lead to lots of good general knowledge in other parts of the platform. I think it’s important that gain both types of skill, as you work through your career and grow your abilities. I find that there is always more to learn, especially in technology. Even in areas that I think I have deep knowledge about, I continually add to, and refine, my skills.

    Brent Ozar had an interesting post on this topic, looking at the maturity of a particular subject area. He noted the more mature some area, the more you might learn towards specialization. I think that’s good advice for your career, though I do think that if you find a deep passion, don’t be afraid to specialize in some area. While you might not get a monetary compensation directly back for the effort put in, I think the reward of learning something you love, and achieving expertise is extremely valuable and important to you as an individual.

    As we move through our careers in computing, we find there are always new technologies, techniques, and architectures that we will need to understand. We need to grow our knowledge. In many cases we see older ideas re-packaged and updated for the modern times, but that doesn’t mean skills we can rely on prior knowledge. Instead, you will need to update your skills.

    I also think more and more we will be driven to give guidance to others as much as use specific expertise to solve problems. With that in mind, I think whether you choose to be a specialist of generalist, you’ll need to continue to learn more about both implementing technology, as well as communicating your recommendations.

    Steve Jones

    The Voice of the DBA Podcast

    Listen to the MP3 Audio ( 2.8MB) podcast or subscribe to the feed at iTunes and LibSyn.