Amazon Managed Workflows for Apache Airflow Reviews

Our use case for Amazon Managed Workflows for Apache Airflow is more focused on operational workflow. We created an operational workflow for one of the related projects, which is purely internal. This means the solution needs a lot of files and requires data from different sources to load into the system, create an index, gather status, trigger alerts, and monitoring. So we use Amazon Managed Workflows for Apache Airflow for this tool and solution.

Amazon Managed Workflows for Apache Airflow is easy to use and offers great scalability. The more we include and extend our capabilities, we can easily scale up according to our needs. When we compare it to other tools in the market, such as Control-M and other technologies, Airflow offers a lot of flexibility. Using this tool minimizes costs and maximizes efficiency.

Customization in general depends on the use cases that we are trying to address, but I don't have a specific thing right now.

I will get some insight from my colleague, and perhaps we can talk about it next week regarding some minimal improvements or issues with the product.

I have been working with Amazon Managed Workflows for Apache Airflow for almost one and a half years, but in fact, my team members are using it to create solutions and create an orchestration tool for one of the projects.

We adhere to their SLA, which is around 98%. From that perspective, it is supposed to be stable, and it is stable anyway.

To be very honest, technical support from AWS is extremely expensive.

We try to involve the professional services from AWS, and we feel that it's quite costly.

If I were to rate the quality of support and response time from 1 to 10, I would give them eight out of ten.

Positive

I was extensively using Control-M, which is another orchestration tool. I usually compare with Control-M, but I don't say that Control-M can replace Amazon Managed Workflows for Apache Airflow or vice versa. Both address different needs. While we can achieve the same methodology and processes with Control-M, we also use Amazon Managed Workflows for Apache Airflow. When I compare Control-M and Amazon Managed Workflows for Apache Airflow, I find that Airflow is pretty much decent enough.

Overall, it should be easy to install Amazon Managed Workflows for Apache Airflow. The reason is that the Control-M product we used last time was purely on-prem, and we needed to maintain an admin server and workflow server, making it a very hectic and tedious process. In contrast, Amazon Managed Workflows for Apache Airflow operates in a purely cloud mechanism, so we don't need to think too much about maintaining the infrastructure and installation.

For Amazon Managed Workflows for Apache Airflow, it is supposed to be easy, and my tech team can take a deep dive on it.

When we discuss return on investment, I go back to the same comparison. When I use Control-M jobs, we have almost around 700 jobs that need to be created, and we also must connect with different source systems. When I compare Amazon Managed Workflows for Apache Airflow with Control-M jobs, I find that the cost we are putting on the resource side is a bit more expensive with Control-M. From a business point of view, using Amazon Managed Workflows for Apache Airflow allows us to optimize the cost by at least 12% annually in OPEX costs.

Custom Generalizations do not directly correlate with specific sections conventional to reviews. Less clarity on direct IMS or broader tech engagement could stem from extensive teamwork or diverse project priorities shared among colleagues. On a scale of one to ten, I rate Amazon Managed Workflows for Apache Airflow a seven out of ten.

We encounter challenges transitioning from a graphical user interface like Control M to a code-based approach.

It does not have a user interface for building or configuring schedules, which makes it difficult for application support teams to add ad-hoc jobs without relying on the development team.

I have been familiar with Amazon Managed Workflows for Apache Airflow for six to seven months.

Based on the documentation, scaling takes about five minutes, but some parameters can be adjusted to reduce this time. Amazon promises that the service is infinitely scalable, but this needs testing.

I rate the scalability a seven.

We have contacted Amazon's technical support with questions regarding best practices. The team is helpful.

Neutral

We chose Amazon Managed Workflows because it is an AWS-managed service, so we do not need to involve another vendor.

The deployment process could have been more straightforward. It involved enabling the service and choosing between small, medium, or large configurations, followed by reading the documentation to determine the best fit for our use case. However, the documentation provided is good. The installation of the service itself takes only a few minutes.

I rate the product pricing a five. It is not an inexpensive solution but it is less expensive than other options like Control M.

One of the product's main strengths is that it is well-suited for a DevOps pattern, allowing us to automate our CI/CD pipeline. The scheduling as code feature works efficiently to promote changes to higher environments and manage multiple version changes during development.

I rate it a six out of ten.