Press "Enter" to skip to content

Category: ETL / ELT

Modern Data Warehousing with Data Lake Storage and Azure Data Factory

Published 2024-05-21 by Kevin Feasel

Josephine Bush continues a series on modern data warehousing:

In today’s data-driven world, having the right tools to manage and process large datasets is crucial. That’s where Azure Data Lake Storage (ADLS) and Azure Data Factory (ADF) come in handy, making it easier than ever to store and transform your data. In this post, I’ll show you how to set up ADLS to store your Parquet files and configure ADF to manage your data flows efficiently.

Read on for an overview of both technologies.

Comments closed

Creating Orchestrators in Azure Data Factory

Published 2024-05-15 by Kevin Feasel

Martin Schoombee continues a series on building an orchestration framework in Azure Data Factory:

The orchestration layer of the framework is where all the magic happens. It facilitates the execution of processes and/or tasks as defined in the metadata, and needs to do it both seamlessly and efficiently. Ideally you would want to deploy this layer only once, and never have to touch it again. And it is really with that in mind that I designed this layer…to function independently and with minimal dependencies in both directions.

I would have loved for this layer to consist of only one pipeline but there are some nuances in Data Factory that make it impossible, the primary nuance being that you cannot nest ForEach activities. As a result, this layer contains three pipelines that will be covered by the sections below in more detail.

Read on to see what those three pipelines are.

Comments closed

Building Workers in Azure Data Factory

Published 2024-05-09 by Kevin Feasel

Martin Schoombee continues a series on orchestration in Azure Data Factory:

We’re finally ready to dive into the Data Factory components that form part of the framework, and we’re going to work our way from the bottom up. To paraphrase the previous blog post, worker pipelines perform the actual work of either moving data (from source to staging) or executing a stored procedure that will load a dimension/fact table.

Although worker pipelines can contain any number of tasks you may need, my worker pipelines that move data from a source system into the staging area follow a similar pattern with at least the following activities:

Click through for that list, as well as more information.

Comments closed

Editing the JSON of a Microsoft Fabric Pipeline

Published 2024-05-09 by Kevin Feasel

Dennes Torres makes a change:

A Fabric Pipeline uses JSON as source code. They are also saved in repositories as JSON.

We first idea we get is editing the pipeline in JSON format. We can copy the JSON and create new pipelines with small variations, making changes directly on the JSON.

However, at first sight we get disappointed, because the pipeline doesn’t allow the JSON to be edited. We have the option to view the JSON, but nothing else.

Read on to see how to tell the Fabric pipeline who’s boss.

Comments closed

Building Metadata-Driven Pipelines in Microsoft Fabric

Published 2024-05-07 by Kevin Feasel

Koen Verbeeck lays out a process:

The goal of metadata driven code is that you build something only once. You need to extract from relational databases? You build one pipeline that can connect to a relational source, and you parameterize everything (server name, database name, source schema, source table, destination server name, destination table et cetera). Once this parameterized piece of code is ready, all you must do is enter metadata about the sources you want to extract. If at a later point an additional relational source needs to be extracted, you don’t need to create a brand-new pipeline. All you need to do is enter a new line of data in your metadata repository.

Aside from speeding up development – after you’ve made the initial effort of creating your metadata driven pipeline – is that everything is consistent. You tackle a certain pattern always in the same way. If there’s a bug, you need to fix it in one single location.

Read on to see how this works. The idea is certainly not new, as Koen mentions, but there are some specific factors that come into play for Microsoft Fabric pipelines.

Comments closed

Three Layers of Azure Data Factory Framework Components

Published 2024-04-24 by Kevin Feasel

Martin Schoombee continues a series on orchestration in Azure Data Factory:

Before we dive into the details of the Data Factory pipelines, it is worth explaining the conceptual structure of my framework and its components. How it all fits together is important, and after reading the post on the metadata as well the pieces of the puzzle will hopefully start falling into place.

When I started thinking about what I’d like the framework to do, three conceptual layers started to emerge and we’ll review them from the bottom up:

Click through for the description of each layer.

Comments closed

Metadata Tables and Azure Data Factory

Published 2024-04-17 by Kevin Feasel

Martin Schoombee brings back metadata tables:

The metadata that drives the execution within a framework is probably the most critical part. Going back to our analogy of building a house, the metadata would be the foundation. It is here where you are going to make some architectural decisions outside of which the framework cannot operate.

One such decision is how configurable or flexible you’d like the framework to be. In other words, how many attributes would you like to be dynamic and/or have the option to change during execution. It seems like an easy choice, and most engineers would lean towards “everything” or “as much as possible” as an answer. In reality however, the trade-off is complexity and the more dynamic you make the framework the more complicated it becomes. And you pay for the complexity later when you need to maintain or add new functionality to it.

Read on to see how it all fits together.

Comments closed

Tips for Configuring Alerts for Azure Data Factory

Published 2024-04-16 by Kevin Feasel

Teo Lachev shares some advice:

Alerting is an important monitoring task for any ETL process. Azure Data Factory can integrate with a generic Azure event framework (Azure Monitor) which makes it somewhat unintuitive for ETL monitoring. You can set up and change the alerts using the ADF Monitoring hub.

Read on for five pieces of advice, in particular, covering how to set up one of these alerts.

Comments closed

A Scaffolding Design Pattern for Microsoft Fabric Pipelines

Published 2024-04-15 by Kevin Feasel

Andy Leonard shares some thoughts on design:

When assigned a project, it’s tempting – and dangerous – to Just Start Coding. If you suffer from the urge to develop first and design later, you are not alone (there’s at least one other developer like you and he’s typing this post). Do yourself a favor and…

Read on for more information on Andy’s design-first mentality and a sample of how you might lay out that initial design.

Comments closed

The Importance of Orchestration in E(L)TL Processes

Published 2024-04-10 by Kevin Feasel

Martin Schoombee begins a new series:

In the context of what we’re talking about throughout this series – facilitating the execution of an ETL process in a platform like Azure Data Factory – orchestration means that we’re using the ETL tool primarily for the “E” (Extract) part of the process. In addition to that, most people I know would also use the ETL tool to facilitate the workflow, in other words the order of execution and any constraints that go along with that.

In what I’d like to call the “traditional” approach for lack of a better term, all parts of the ETL process are performed natively by the tool (image below), using whatever built-in tasks are available and of course accounting for any nuances. With this approach, transformations are typically performed in transit and in memory.

Read on to see how the Orchestration approach differs from the traditional ETL approach.

Comments closed