A Fine Slice Of SQL Server
Gilbert Quevauvilliers does a bit of writing:
In this blog post I am going to explain how to use a Python Notebook using the Semantic Link module, to run a DAX query and write the output to a Lakehouse table.
I will show you how to install a Python library and then use it within my python notebook.
Read on for a quick primer on Semantic Link Labs, followed by the meat of the article.
Leave a CommentChad Callihan shares a pair of tips:
These aren’t the most technical SQL Server Agent job topics, but two that came to mind were managing who jobs belong to and making sure the right steps are followed when jobs are removed.
Click through for more thoughts on both of these topics.
Leave a CommentHaripriya Naidu is trying to slam a lot of transactions through the same door:
When operations wait to acquire a latch on a page, you’ll see a wait type called
PAGELATCH. A latch is a lightweight lock.PAGELATCHwaits typically occur in TempDB on pages like PFS, GAM, SGAM, or system object pages.Normally, you won’t see
PAGELATCHwaits in user databases because user objects don’t usually experience the same level of concurrent inserts/updates/deletes as temp tables do.But, there is one case where this can happen:
When many concurrent transactions try to insert into the last page of a table, they all compete for a latch on that page. This results in last page insert contention.
Read on to see when this happens, as well as a demonstration of it. Haripriya then uses a bit of functionality that is available in recent versions of SQL Server to resolve the issue.
Leave a CommentDavid Wiseman shows off an open-source product:
For T-SQL Tuesday #186, Andy Levy asks, “How do you manage and/or monitor your SQL Server Agent jobs?”
This is a great opportunity for me to discuss how DBA Dash can help monitor SQL Agent jobs. DBA Dash is a free and open-source monitoring tool for SQL Server, created by me. It’s used to monitor thousands of SQL Server instances within Trimble alone, and it’s gaining popularity in the SQL Server community.
Read on to see how the product can help if you have a series of SQL Agent jobs.
Leave a CommentWhen I presented “Choosing Between Import Mode, Direct Lake, and Composite Models” at Fabric Conf 2025 in Las Vegas, the room overflowed, and the session was not recorded. I promised to publish the material once the new Direct Lake + Import composite model became available. This post follows the structure of that (now re‑recorded) session.
I prepared a recap for this blog post, but I suggest you watch the full video!
Check out the video and Marco’s guidance.
Leave a CommentI used to work in a fairly large enterprise (5,000+ people, 500+ production SQL instances) with a small staff. It was 2-3 of us to manage all these systems, as well as respond to questions/queries/issues with dev/test systems. As a result, we depended heavily on SQL Agent.
We decided on a few principles which helped us manage jobs, with a (slow) refactoring of the existing jobs people randomly created with no standards. A few of the things we did are listed below. This isn’t exhaustive, but these are the main things I remember.
Read on for Steve’s list.
Leave a CommentTodd Kleinhans does not believe that green is good:
Far too many times, I have seen DBA(s) and others have this false sense that if the Agent run status shows green, then everything must be ok.
Click through for a funny story about a gas station robbery and examples of how a SQL Agent job can report success but actually fail. You also see this a lot with replication or tasks that are asynchronous in nature: the task is reporting that we successfully started whatever operation, but that doesn’t mean the operation itself succeeded.
Leave a CommentIn this video, I show how we can use the APPLY operator to operate on ad hoc functions. That leads to a powerful use case: pre-aggregating data.
Every once in a while, this tip will save a considerable amount of CPU time and database effort.
Leave a CommentDeborah Melkin has three laws for robots, or at least SQL Agent jobs:
How do I manage this all?
I have scripts to create and modify SQL Agent jobs.
And those scripts get checked into source control.
And those scripts get added to the deploy process.
And most importantly, those scripts get tested in all environments from development through production.
Read on for more context around this.
Leave a CommentStephen Planck adds one more layer of complexity:
SQL Server’s database-snapshot feature is a wonderfully simple tool: at the instant you create the snapshot, every page in the database is marked “copy-on-write.” Nothing is copied across the wire, no blocking locks appear, and the snapshot opens immediately as a read-only database on the local replica. Queries against the snapshot see the world exactly as it looked at that moment while the live workload keeps changing pages in the primary data files. Because snapshots live only in sparse files on the server that owns them, they are not a replacement for backups—but they are perfect for ad-hoc reporting, quick “before-and-after” comparisons, or a safety net when you want an easy way to back out a risky change that should finish within minutes or hours.
But read on to see how they interact with high-availability features such as transactional replication and availability groups.
Leave a Comment