A Fine Slice Of SQL Server
Sandeep Pawar keeps things up to date:
It’s always a good idea to use the latest GA runtime for the default Spark pool in Fabric workspaces. Unless you change it manually, the workspace will always use the previously set runtime even if a new version is available. To help identify the runtime that workspaces are using and to upgrade multiple workspaces at once, use the code below, powered by Semantic Link.
Read on to see how you can do it using a bit of Python scripting.
Comments closedLouis Davidson looks back on the past:
I will admit that the feature I will mention probably won’t be relevant to you, as it has been nearly 30 years since this feature was introduced. But since this is the most nostalgic time of the year, I want to go back, through the mists of times, to a time where we hadn’t even quite begun to fear the changing of the centuries. To a time when we a single CPU was all we had even thought of, computer screens were often colorless, a few megabytes of RAM was all that was needed. It was a simpler time, though to be fair, it was also a time before we had sports on our phones.
Louis goes on to reminisce about an implementation of database quality guarantees in a world without foreign key constraints. Which is usually about the time I rag on MySQL for its insistence that foreign keys weren’t important to the product until relatively late in its existence (version 3.23, released in 2000).
Comments closedBrent Ozar lays out an argument:
So you’re writing T-SQL code and you’ve decided that you want to send emails to customers, employees, or the general public. Perhaps you need to send order updates or low stock notifications.
Stop right there. You don’t really wanna do that with SQL Server. Let’s talk about why.
Read on for good reasons why. The bit of fuzziness here is “employees,” as Database Mail is quite useful for DBAs or individuals responsible for the upkeep of database-related operations. But the further you move away from “People whose job it is to keep the SQL Server instance up and running,” the less viable Database Mail is as a product, for the exact reasons Brent mentions.
Of course, you could always do what I did at one consulting engagement and set up sending Slack messages via CLR. I’m not saying it was a good idea, but it was pretty cool that it worked as well as it did.
Comments closedVitaly Bruk deals with a CPU issue:
Today, I investigated an interesting issue.
One of my clients called me and complained about high CPU on his server. Server with 0 load at this time frame.
Using my favorite “Activity” script and the sys.dm_os_ring_buffers DMV, I clearly saw that the CPU is being used by a non-SQL server service. Next, I’ve asked to connect SQL Server machine and opened a task manager.
Surprise, surprise! The CPU was being used by the SQL Server process! Well, kind of…
The moral of the story is to keep your systems patched.
Comments closedSQL Server Reporting Services (SSRS) uses the object type of a data source as the basic unit of connecting a data source to a report. (See this tip – SQL Server Reporting Services SSRS 2017 Data Sources) A data source is the connection details that allow a report designer and report consumer to define where the data lives and includes items like a table in SQL Server or an API. Continuing this concept, the dataset is the basic object that defines the structure of the data to be loaded and establishes a collection of fields.
Read on to dive into the properties of an SSRS dataset.
Comments closedKai Waehner lays out the state of things:
Data streaming is a new software category. It has grown from niche adoption to becoming a fundamental part of modern data architecture, leveraging open source technologies like Apache Kafka and Flink. With real-time data processing transforming industries, the ecosystem of tools, platforms, and cloud services has evolved significantly. This blog post explores the data streaming landscape of 2025, analyzing key players, trends, and market dynamics shaping this space.
It’s always important to keep the writer’s bias in mind when reading these articles (and we all have biases, whether or not we admit to them). With that preparatory throat-clearing out of the way, Kai does an excellent job laying out the players, the criteria he uses for analysis, and the current state of the field.
Comments closedTomaz Kastrun takes a peek at the Azure AI Services Language and Translation UI:
In Language + Translation UI you will give you capabilities to integrate the natural languages into your apps, bots and other services. With Language + Transport you can redact sensitive data, segment your text, and many more.
Click through for a demonstration.
Comments closedHugo Kornelis looks back at some nice additions to SQL Server:
It’s December. The last month of the year, and hence also the last T-SQL Tuesday of the year. Edition 181 is hosted by Kevin Chant. His chosen topic for this episode is to have us talk about which SQL Server announcement excited us the most.
This posed a bit of a problem for me. The only truthful answer here is that I have never been as excited about an announced new feature as when the Query Store was announced. But I also realized that I don’t have much to write about the Query Store, except how awesome it is. Okay, I could go on a full explanation, but many others do that much better, and have already done so. So that doesn’t leave much for me.
And that’s why I’ll also include my second favorite announcement.
Read on for a little bit around Query Store, followed up with coverage of several interesting additions to SQL Server over the years.
Comments closedIn this video, I show how to back up and restore SQL Server databases in piecemeal form, using filegroups to manage read-only versus read-write data, and bringing the most important data back online sooner in a recovery scenario.
I found the process to be a bit trickier than I had first expected, so I’m hoping this video has enough legs to prevent others from running into some of the problems I experienced.
Comments closedRob Farley cashed in all of his chips and got something nice out of it:
I had written to Santa (okay, the product group at Microsoft) to ask for this. It wasn’t quite “If I’m on the nice list, what I’d like is a unicorn train set”, but more like “If it’s not too much trouble, this thing here would be amazing…”
The thing was the ability to see the “number of rows read” by an Index Seek operation in a SQL plan containing the “Actuals” (post-execution).
Read on for information about why this is so useful.
Comments closed