by yohamta on 7/18/22, 2:46 AM with 2 comments
Why not Airflow? Airflow and other similar tools are powerful and valuable, but in most cases, they require writing code to manage workflows.
Our ETL pipeline is already hundreds of thousands of lines of complex code in Perl and shell scripts. Adding another layer of Python on top of this would make it difficult to maintain. Instead, we needed a more lightweight solution.
So we developed Dagu, which requires no coding, and is easy-to-use and self-contained, making it ideal for smaller projects with fewer people.
by theamk on 7/18/22, 3:57 AM
If the goal was to eliminate project-specific DAG code in programming languages, have you considered writing an YAML adapter for an existing tool? For example, an "generic" Airflow pipeline which reads YAML files and generates tasks for each clause. This would have much less code than writing scheduler from scratch, and none of it would be project-specific.