Shoaib
Shoaib I’m curious about how things work — and how data can improve the way we solve problems and make decisions. Everything I share is part of that exploration, created not just to inform, but to inspire smarter ways of thinking, building, and solving.

Configuring Azure Data Flow Error Handling

In this task, we are going to configure Azure Data Flow Error Handling.

We have 2 containers in our storage account

  • bad-records
  • demo-data

Inside demo-data, we have a csv file “People.csv”

first_name,last_name,age
Lenny,Kravitz,46
Steven,Segal,55
Alice,Jones,23
Paula,Stevens,45
Mandy,Patinkin,63
Randall,Flagg,32
Steven,Segal,55
Paula,Stevens,45
Mandy,Patinkin,62

Inside out sql database, we have created a table called People.

Notice the first_name has 5 character limit. This is created to do this task of error handling.

1
2
3
4
5
6
CREATE TABLE People
(
first_name varchar(5),
last_name varchar(25),
age int
)

Data factory steps

1. Create linked services

Create linked service to blob storage for our source

Create linked service to SQL database for our sink

2. Create datasets

1 dataset for People.csv in blob storage

1 dataset for bad records in blob storage

1 dataset for People table in SQL database

3. Create new dataflow

Choose source

Use conditional split to split good records (length(first_name)<=5) and bad records

Choose sink for good records -> sql database table

Choose sink for bad records -> bad-records container

  • choose output as single file and give the file name

alt text

4. Create pipeline and run

Create a pipeline, and publish all

Click trigger now to run the pipeline

5. Check the SQL database and container

You should now see the table in SQL database populated as per the conditional logic

You should also see the bad-records container, a new file is created based on conditional logic.

You should see a csv file with first_name > 5 characters.