Press "Enter" to skip to content

Author: Kevin Feasel

Kafka Internals: Handling a Producer Request

Published 2024-09-26 by Kevin Feasel

Danica Fine continues a series on Kafka internals:

Welcome to the second installment of our blog series to understand the inner workings of the beautiful black box that is Apache Kafka®. 

We’re diving headfirst into Kafka to see how we actually interact with the cluster through producers and consumers. Along the way, we explore the configurations that affect each step of this epic journey and the metrics that we can use to more effectively monitor the process. 

In the last blog, we explored what the Kafka producer client does behind the scenes each time we call producer.send() (or similar, depending on your language of choice). In this post, we follow our brave hero, a well-formed produce request, that’s on its way to the broker to be processed and have its data stored on the cluster.

Click through to learn more about how it all works.

Comments closed

Copying Content between Databricks Unity Catalogs

Published 2024-09-26 by Kevin Feasel

Meagan Longoria busts out the photocopier:

I had a couple of clients who were moving content from development catalogs to production catalogs for the first time. They wanted to copy the schema and data from tables, views, and volumes.

So I wrote a python notebook to handle this task. It creates the objects in the new catalog and then changes the owner to a selected user/group. You can find the full notebook on GitHub.

You can check out the notebook, and Meagan also has explanatory notes covering the process.

Comments closed

Cardinality Estimation Changes Post-SQL Server 2014

Published 2024-09-26 by Kevin Feasel

Brent Ozar reminds us that small changes happen:

About 10 years ago, Microsoft made changes to the Cardinality Estimator (CE) which caused some problems for SQL Server upgrades. When folks upgraded to SQL Server 2014, they also casually switched their databases’ compatibility level to the latest version, because for years that hadn’t really affected query plans. They just figured they wanted the “latest and greatest” compat level, without regard to the effects. That backfired badly when they suddenly got 2014’s Cardinality Estimation changes.

So for several years, whenever someone upgraded from older versions, and they complained about performance, the stock community answer was, “Change your compatibility level back to what it used to be.” In many cases, that just solved the problems outright, leading to blog posts like this and this.

Even today on SQL Server 2019 & 2022, this advice is still relevant! If you mess around with compatibility levels, you can absolutely change cardinality estimations in ways you didn’t expect.

Read on for more information, including examples where cardinality estimation has improved with SQL Server 2022 and where it has gotten worse.

Comments closed

Tracking Microsoft Fabric Notebook Progress

Published 2024-09-26 by Kevin Feasel

Gilbert Quevauvilliers asks are we there yet? are we there yet?

How to view or track the progress of Notebook while it is running in Microsoft Fabric

I was recently working with a Notebook in Microsoft Fabric that was started via a Data Pipeline.

The challenge I had was that I had no idea how far the notebook had gone (as there were quite a lot of cells in this particular notebook).

In this blog post I am going to show you how I can use Microsoft Fabric to identify exactly which cell my notebook is currently on.

Click through for the answer. And so help me, if you ask that question one more time, I’m turning this thing around and we’re going back home.

Comments closed

Online Upgrade of Logical and Physical Replication Notes in Postgres

Published 2024-09-26 by Kevin Feasel

Amit Kapila notes an improvement in replication capabilities:

In my last blog post, Evolution of Logical Replication, I mentioned the future development of a feature to allow “upgrades of logical replication nodes.” The upcoming release of PostgreSQL 17 includes this feature. Previously, after major version upgrades (via pg_upgrade), users couldn’t immediately connect and write data to logical replication nodes. This was because the slots were lost during upgrades, preventing replication from continuing. As a result, new writes wouldn’t get replicated, causing data on both nodes to become out of sync. As explained in this blog post, users had to block applications from writing until the replication setup was re-enabled after the upgrade.

Read on to see how PostgreSQL 17 changes things.

Comments closed

Debugging a DAX Measure with DAX Query View

Published 2024-09-26 by Kevin Feasel

Dennes Torres sorts out an issue:

DAX Query View was created some time ago and it was a great tool to run DAX inside Power BI environment.

However, testing a measure in DAX Query View may not be something so obvious. While the measure in a report obeys to the filter context and most of times results in a single value, the filter context doesn’t exist in Query View and the execution needs to return a table.

In this way, a simple copy/past of the measure in DAX Query View will not run. The measure needs to be adapted to be tested.

Let’s analyze one sample case to discover how this happens.

Click through for the answer.

Comments closed

Azure Database for PostgreSQL Backup and Restoration

Published 2024-09-26 by Kevin Feasel

Josephine Bush makes and restores a backup:

I wanted to test a migration of Azure PostgreSQL from simple to flexible. I wanted a simple db to restore onto my single server. I may not have needed to do this, but I was then curious how one would go about dumping and restoring on Azure Postgres in general. I plan to use the Azure migration service, but since I started this dump/restore exercise, I thought I would document it here.

Read on for the process.

Comments closed

Tracking Python Packages in Snowflake

Published 2024-09-25 by Kevin Feasel

Kevin Wilkie takes a peek:

When working with one of the many modern computer languages that use libraries, one of the many things to be aware of – as a developer – is the version of the libraries available for your usage.

Since there are multiple languages in Snowflake that use libraries, let’s go over how to check out the versions that come installed and how to install one yourself.

Read on for those answers. Well, one answer and one conundrum.

Comments closed

Simple Outlier Detection and Removal in R

Published 2024-09-25 by Kevin Feasel

Steven Sanderson looks for oddities:

Outliers can significantly skew your data analysis results, leading to inaccurate conclusions. For R programmers, effectively identifying and removing outliers is crucial for maintaining data integrity. This guide will walk you through various methods to handle outliers in R, focusing on multiple columns, using a synthetic dataset for demonstration.

The techniques Steven uses are perfectly reasonable (though I like to use MAD from the median rather than standard deviations from the mean because MAD from the median doesn’t suffer from the sorts of endogeneity problems standard deviation does in a dynamic process). My primary warning would be to keep outliers in a dataset unless you know why you’re removing them. If you know the values were impossible or wrong—for example, a person who works 500 hours a week—that’s one thing. But sometimes, you get exceptional values out of an ordinary process, and those values are just as real as any other. I might have had a sequence in which I flipped a fair coin and it landed on heads 10 times in a row. It’s statistically very uncommon, but that doesn’t mean you can ignore it as a possibility or pretend it didn’t happen.

Comments closed

Data Ingestion with Microsoft Fabric Copy Jobs

Published 2024-09-25 by Kevin Feasel

Reitse Eskens spends a bunch of time at the copier:

The copy job is essentially an abstraction of a pipeline reading data from the source system and writing the data into either a Lakehouse or a Warehouse. It really is ingesting data and nothing else. In my opinion that what copy data flows are meant to do and are very good at too.

The big challenge we all keep facing is how to create incremental loads. We have to build some sort of metadata database where we keep the latest ID, data or other column we use to discern the increment on. In our flow, we need to get that value, compare it against the source system and get the differences. The biggest task is to find out if records are deleted.

With the Copy Job, a large part of this task is taken out of your hands. The Copy Job has a configuration GUI (or wizard) that helps you out quite quickly. So let’s not waste anymore characters and dig in!

Read on to see how it works and its capabilities and limitations. The key question, as always, is whether your workload fits into the wheelhouse. If so, this sounds really useful. If not, it’s a proper struggle.

Comments closed