Amazon MWAA

Airflow is a platform created by the community to programmatically author, schedule and monitor workflows. Airflow has a modular architecture and uses a message queue to orchestrate an arbitrary number of workers. Airflow is ready to scale to infinity. Airflow pipelines are defined in Python, allowing for dynamic pipeline generation. This allows for writing code that instantiates pipelines dynamically. Easily define your own operators and extend libraries to fit the level of abstraction that suits your environment. Airflow pipelines are lean and explicit. Parametrization is built into its core using the powerful Jinja templating engine. No more command-line or XML black-magic! Use standard Python features to create your workflows, including date time formats for scheduling and loops to dynamically generate tasks. This allows you to maintain full flexibility when building your workflows.

You select the size of the cluster, node capacity, and a set of services, and Yandex Data Proc automatically creates and configures Spark and Hadoop clusters and other components. Collaborate by using Zeppelin notebooks and other web apps via a UI proxy. You get full control of your cluster with root permissions for each VM. Install your own applications and libraries on running clusters without having to restart them. Yandex Data Proc uses instance groups to automatically increase or decrease computing resources of compute subclusters based on CPU usage indicators. Data Proc allows you to create managed Hive clusters, which can reduce the probability of failures and losses caused by metadata unavailability. Save time on building ETL pipelines and pipelines for training and developing models, as well as describing other iterative tasks. The Data Proc operator is already built into Apache Airflow.