I am currently working on a very interesting ETL project using Azure to transform my data manually. However, transforming data manually can be exhausting and lengthy when I start having several source files to process. My pipeline is working fine for now because I have only a few files to transform but what if I have thousands of excel files?
So what I want to achieve is that I want to extend the project and extract the excel files that are coming from Email using the logic app then apply ETL directly on top of them. Is there any way I can automate ETL in Azure. Can I do ETL without modifying the pipeline for a different type of data manually? How can I make my pipeline flexible to be able to handle data transformation for various types of source data?
Thank you in advance for your help.
Can I do ETL without modifying the pipeline for a different type of
data manually?
According to your description, i suppose that you already knew the ADF connector is supported in the Logic App. You could execute ADF pipeline in the Logic App flow and even pass parameters into ADF pipeline.
Normally, the source and sink service should be fixed in one copy activity, but you could define dynamic file path in the datasets. So you don't need to create multiple copy activities.
If the data types are different, you could try to pass the parameter from Logic App into ADF. Then before the data transmission, you could use Switch activity to route the transmission into different branches.
Related
I have a folder of pipelines, and I want to execute the pipelines inside the folder using a single pipeline. There will be times when there will be another pipeline added to the folder, so creating a pipeline filled with Execute Pipelines is not an option (well, it is the current method, but it's not very "automate-y" and adding another Execute Pipeline whenever a new pipeline is added is, as you can imagine, a pain). I thought of the ForEach Activity, but I don't know what the approach is.
I have not tried this approach but I think we can use the
ADF RestAPI to get all the details of the pipelines which needs to be executed. Since the response is in JSON you can write it back to temp blob and add filter and focus on what you need .
https://learn.microsoft.com/en-us/rest/api/datafactory/pipelines/list-by-factory?tabs=HTTP
You can use the Create RUN API to trigger the pipeline .
https://learn.microsoft.com/en-us/rest/api/datafactory/pipelines/create-run?tabs=HTTP
As Joel called out , if different pipeline has different count of paramter , it will be little messy to maintain .
Folders are really just organizational structures for the code assets that describe pipelines (same for Datasets and Data Flows), they have no real substance or purpose inside the executing environment. This is why pipeline names have to be globally unique rather than unique to their containing folder.
Another problem you are going to face is that the "Execute Pipeline" activity is not very dynamic. The pipeline name has to be known as design time, and while parameter values are dynamic, the parameter names are not. For these reasons, you can't have a foreach loop that dynamically executes child pipelines.
If I were tackling this problem, it would be through an external pipeline management system that you would have to build yourself. This is not trivial, and in your case would have additional challenges because of the folder level focus.
I’m using mosaic decisions data flow feature to read a file from Azure blob, do a few transformations and write that data back to Azure. It worked fine except that in the output file path I have given, it created a folder and I can see many files with some strange “part-000” etc in their names. What I need is a single file in that output location – Not many. Is there a way around this?
Mosaic-Decisions uses apache spark as its backend execution engine. In Spark, the dataframe read is split into multiple partitions and these partitions are written to the output location in parallel. That's the reason it creates multiple files at the target location with "part-0000", "part-0001" etc. (part here represents partition).
The workaround on this is to check "combine-output-files-into-one" in writer node. This will combine all of the part files into one big file. But use this with caution and only if you really need a single file - as this will come with a performance tradeoff.
We have several large CSV files in Azure Data Lake Store that were created using the Append method of the .NET API. Recently, we switched over to ConcurrentAppend for performance reasons. Since ConcurrentAppend and Append cannot be used interchangeably, the switch required us to create a new folder structure for the files, to make sure that the ConcurrentAppend would never hit any files created using Append.
However, our downstream application needs to load all data, both from before and after the switch. Instead of changing our application, we wanted to join the files (using the PowerShell SDK Join-AzureRmDataLakeStoreItem cmdlet), but the documentation does not specify whether files joined this way can be written to by ConcurrentAppend after the join. I suspect that we will face issues, since we are going to join files created by both methods (maybe it's not even possible to do the join?)
So my questions are as follows:
Can ConcurrentAppend write to a file that has been joined using Join-AzureRmDataLakeStoreItem, even if one or more of the source files have been created using Append?
If not, we will use U-SQL to combine the files, but can ConcurrentAppend write to a file that has been outputted from a U-SQL job?
If not, do we have any other options than executing a local script (using the .NET API for example), which will read all files, and write a new set of files back to the lake using only ConcurrentAppend?
Cost is a concern, which is why we prefer to use the PowerShell cmdlet if possible, and would like to avoid the last option.
At present after the join operation, no append operations can be executed on the file. We are currently working on a feature to remove this limitation. However, at present after concatenating files, the appends will not work.
I've been looking around for a lightweight, scaleable solution to enrich a CSV file with additional metadata from a database. Each line in the CSV represents a data item and the columns the metadata belonging to that item.
Basically I have a CSV extract and I need to add additional metadata from a database. The metadata can be accessed via ODBC or REST API call.
I have a number of options in my head but I'm looking for other ideas. My options are as follows:
Import the CSV into a database table, apply the additional metadata with sql UPDATE statements by finding the necessary metadata with SELECT statements, and then export the data back into CSV format. For this solution I was thinking to use an ETL tool which may be a bit heavyweight to tackle this problem.
I also thought about a NodeJS based solution where I read the CSV in, call web service to get the metadata and write back the data into the CSV file. The CSV can be however quite large with potentially tens of thousands of rows so this could be heavy on memory or in case of line-by-line processing not very performant.
If you have a better solution in mind, please post. Many thanks.
I think you've come up with a couple of pretty good ideas here already.
Running with your first suggestion using an ETL tool to enrich your CSV files, you should check out https://github.com/streamsets/datacollector
It's a continuous ingestion approach, so you could even monitor a directory of CSV files to load as you get them. While there's no specific functionality yet for doing lookups in a database, its certainly possible in a number of ways (including writing your own custom logic in Java, or a script in python or JavaScript).
*Full disclosure I work on this project.
My team at work is currently looking for a replacement for a rather expensive ETL tool that, at this point, we are using as a glorified scheduler. Any of the integrations offered by the ETL tool we have improved using our own python code, so I really just need its scheduling ability. One option we are looking at is Data Pipeline, which I am currently piloting.
My problem is thus: imagine we have two datasets to load - products and sales. Each of these datasets requires a number of steps to load (get source data, call a python script to transform, load to Redshift). However, product needs to be loaded before sales runs, as we need product cost, etc to calculate margin. Is it possible to have a "master" pipeline in Data Pipeline that calls products first, waits for its successful completion, and then calls sales? If so, how? I'm open to other product suggestions as well if Data Pipeline is not well-suited to this type of workflow. Appreciate the help
I think I can relate to this use case. Any how, Data Pipeline does not do this kind of dependency management on its own. It however can be simulated using file preconditions.
In this example, your child pipelines may depend on a file being present (as a precondition) before starting. A Master pipeline would create trigger files based on some logic executed in its activities. A child pipeline may create other trigger files that will start a subsequent pipeline downstream.
Another solution is to use Simple Workflow product . That has the features you are looking for - but would need custom coding using the Flow SDK.
This is a basic use case of datapipeline and should definitely be possible. You can use their graphical pipeline editor for creating this pipeline. Breaking down the problem:
There are are two datasets:
Product
Sales
Steps to load these datasets:
Get source data: Say from S3. For this, use S3DataNode
Call a python script to transform: Use ShellCommandActivity with staging. Data Pipeline does data staging implicitly for S3DataNodes attached to ShellCommandActivity. You can use them using special env variables provided: Details
Load output to Redshift: Use RedshiftDatabase
You will need to do add above components for each of the dataset you need to work with (product and sales in this case). For easy management, you can run these on an EC2 Instance.
Condition: 'product' needs to be loaded before 'sales' runs
Add dependsOn relationship. Add this field on ShellCommandActivity of Sales that refers to ShellCommandActivity of Product. See dependsOn field in documentation. It says: 'One or more references to other Activities that must reach the FINISHED state before this activity will start'.
Tip: In most cases, you would not want your next day execution to start while previous day execution is still active aka RUNNING. To avoid such a scenario, use 'maxActiveInstances' field and set it to '1'.