I've decided that, in fact, it is time to start moving people off the ancient technology, Profiler. Before, I always said, stay where you're comfortable. However, keeping people comfortable means that they're also going to keep promoting Profiler/Trace to new people on new platforms. That is a real problem. To fix the problem of old school, slow, inferior, methods of data collection, troubleshooting, and consuming metrics, we need to educate people. Extended Events are not simply a replacement for Trace. They're not simply another way to gather query metrics. No, in fact, this is a whole new tool, with new functionality and a very high level of support and engagement from Microsoft. Extended Events are where all new functionality since 2012 provides mechanisms for monitoring behavior. As technologists we should…
I wrote a short blog post about the misperception that Profiler was easier than Extended Events when it came to the core concept of "click, connect, BOOM, too much data". Go read it if you like, but I don't think it's actually an effective argument for how much easier Extended Events is than Profiler. Here, we're going to drill down on that concept in a real way. Let's start with a little clarification. I'm going to be a little lazy with my language. Trace is a scripted capture of events on a server. Profiler is a GUI for consuming a Trace, either live or from a file, and for creating Trace events. However, almost everyone refers to 'Profiler' when they mean either Trace or Profiler. I may do the same…
I know, I know, I'm on #teamexevents so I think that Extended Events can do no wrong, but let's address this thought that Profiler is easier. Now, if we're strictly talking knowledge, sure, if you've got a lot of experience with Profiler/Trace and very little with Extended Events, of course Profiler is easer. However, what I'm told is that Profiler doesn't require very much set up, while Extended Events does. That's just wrong, but let's put it to the test. The Test For the comparison, we're not going to do anything special with either tool. I'm just going to start collecting query data with the fewest possible clicks and/or key strokes. I'm going to use both tools short cuts to make this as fast as possible. The goal is, click,…
If you go through all the stuff I've written about Extended Events, you'll find that I use causality tracking quite a bit. However, I've never just talked about what causality tracking is and why I use it so frequently. Let's fix that issue now. Causality Tracking Causality tracking is quite simple to understand. It's property that you set for a given session. A session, of course, is defined by one or more events and a target. You can define things about a session, like it's name, when you define the session itself. Turning on, or enabling, causality tracking is just a matter of defining that the session will have causality tracking. It looks like this in the GUI: It looks like this in the T-SQL code: CREATE EVENT SESSION QueryBehavior…
The first time you see a new execution plan that you're examining to fix a performance problem, something broken, whatever, you should always start by looking at the first operator. First Operator The first operator is easily discerned (with an exception). It's the very first thing you see in a graphical execution plan, at the top, on the left. It says SELECT in this case: This is regardless of how you capture the execution plan (with an exception). Whether you're looking at an execution plan from the plan cache, Query Store, or through SSMS, the execution plan, regardless of complexity, has this first operator. In this case, it says UPDATE: If you get an execution plan plus runtime metrics (previously referred to as an "actual" execution plan), you'll still see…
A question that I've seen come up frequently just recently is, how to track CPU use over time. Further, like a disk filling up, people want to know how to predict their CPU usage, so that they can easily decide "now is when I upgrade the hardware". Well, the bad news is, that ain't easy. CPU Use Over Time There are a bunch of ways to look at processor usage. The simplest, and probably most common, is to use the Performance Monitor counters such as '% Processor Time'. Query this, you can get an average of the processor usage at a moment in time. Ta-da! Fixed it. I thought you said this was hard Grant. Well, hang on. Are you running on a single processor machine? If so, cool, maybe…
In April, I said I was going to start learning Jupyter Notebooks. It's November. Let's get going with your first Jupyter Notebook. A quick aside before we start. I think one of the huge strengths that is going to come out of these things is as a runbook. You can share a notebook with someone, they can run the queries on it against their own systems and return the book, with the results to you. That's going to be extremely useful as a troubleshooting tool, but has all sorts of other functionality as well. I strongly suggest you start learning these things, as I am. Azure Data Studio There are a number of ways to create and consume Jupyter Notebooks, but I want to focus on the functionality around data…
I consider myself to be the most responsible for making such a huge deal about the differences between what is labeled as an Estimated Plan and an Actual Plan. I walked it back in the second edition of the Execution Plans book. Hugo and I completely debunked the issue in the third edition of the Execution Plans book. That is the one you should all be referencing now. As I like to joke, the guy who wrote the first two editions of the book was an idiot (and lest anyone take offense, let's be clear, I'm the idiot). Now, I'm trying my best to make this whole issue more clear. Let's talk about the "different" plans you can capture in SQL Server. Estimated Plan This is where you have a…
In order to take advantage of R and Python (and Java in SQL Server 2019) directly from your SQL Server scripts, you'll be using the function sp_execute_external_script. When you see this code in use for the first time, it's going to remind you of sp_execute_sql. The very first thing I thought about was, "Oh no. Another SQL Injection vector." I have a little good news and a little bad news. It's Not SQL The first and most important thing to understand is, we're not talking about SQL. Let's start with looking at some code. This is straight from the examples in the Microsoft documentation linked above: DROP PROC IF EXISTS generate_iris_model; GO CREATE PROC generate_iris_model AS BEGIN EXEC sp_execute_external_script @language = N'R' , @script = N' library(e1071); irismodel <-naiveBayes(iris_data[,1:4], iris_data[,5]);…
Not really, but sort of. The beauty of containers, at least in a dev/test environment, is the ability to spin them up while you need them and then throw them away when you're done. Containers give you a bunch of functionality not otherwise available through a VM. However, once you've spun up a container, they're so dull. Why Are Containers Boring Grant? I'm so glad you asked. Last week I was presenting at SQLIntersection (great show, you should consider attending). I was talking about Query Store in SQL Server 2019. One person in the audience asked, "Can Query Store run inside a container?" I responded, "Great question, let's check." I then switched over to VS code to show this: docker run ` --name DemoSharedVol ` -p 1460:1433 ` -e "ACCEPT_EULA=Y"…