We previously used Amazon Kinesis Data Analytics for real-time stream processing. AWS later renamed the service to Amazon Managed Service for Apache Flink, and with the retirement of SQL Applications, we've transitioned our workloads to Apache Flink applications while maintaining the same streaming architecture.

Amazon Kinesis Data Analytics allowed us to build real-time streaming pipelines without managing Apache Flink clusters ourselves. I particularly liked support for event-time processing, stateful streaming, windowed aggregations, checkpointing, and automatic scaling. After migrating to Amazon Managed Service for Apache Flink, these capabilities became even more powerful while maintaining operational simplicity.

The service integrates naturally with Kinesis streams, Kafka, Amazon S3, and CloudWatch. Performance has been excellent for processing millions of streaming events with very low latency. Automatic scaling and managed infrastructure significantly reduced operational overhead compared to self-hosted Flink clusters. Review collected by and hosted on G2.com.

Developing complex streaming applications requires familiarity with Apache Flink concepts, making the learning curve steeper than traditional batch processing. Debugging distributed stream processing jobs can also be challenging. Review collected by and hosted on G2.com.

According to understanding past 2 year I worked on it AWS Analytics offers scalable and flexible solutions for data analysis. Its cost efficient pay as you go model, seamless integration with other AWS services, and managed nature simplify workflows. Real-time capabilities, security features, and machine learning integration enhance its value for various use cases. Review collected by and hosted on G2.com.

According to me in AWS Analytics Challenges with the complexity in tool selection and setup, learning curves for new users, careful cost monitoring, potential service and limited customization. Review collected by and hosted on G2.com.

Easy to setup and integrate with IOT devices and IOT applications with scalable data streaming and processing real-time data analytics with multiple peer devices Review collected by and hosted on G2.com.

At max we can create 50 data streams at a time, if we need more we may need to contact support .. again it is a tedious task. A single shard can ingest upto max 1MB of data or 1000 messages at a time. Review collected by and hosted on G2.com.

I really like to work with the tool and like its capabilities. Easy to deploy even Flink is complex stuff. After understanding the concept, it can be used in a very intense situation. The possibility of the use cases is endless. What I also really like about the product, is that AWS guys are really supportive of this product. Review collected by and hosted on G2.com.

It's not about the product but the available resource that can support you. Since it's a complex tool, and not widely used, it's hard to find reliable support, even in AWS. I believe a lot has to invest in evangelization and learning possibilities of the tool. Review collected by and hosted on G2.com.

Well managed, easy to use. One can learn and use it very easily. Review collected by and hosted on G2.com.

It is a very well-managed service. Not a queue system, so little visibility into "backlog" if there is any. Review collected by and hosted on G2.com.

Apart from its scalability, I would say processing real-time data collected from the web services that we use in our organization. In my personal opinion I have used Kinesis for developing applications for customers who asked about real-time data processing. Review collected by and hosted on G2.com.

My team and myself have been using Kinesis for quite some time and I will be very honest that I have nothing that I can mention about what I dislike about Kinesis. Review collected by and hosted on G2.com.

Mimesis has been working well for us. Initially it took some time to get around this technology but later. After reading in detail could implement it. The examples were limited too at the time of the development. After initial hiccups, the availability and speed has no comparison. We could get data from SAP to caching layer in milliseconds. Review collected by and hosted on G2.com.

AWS Team, you have made this service very daunting at the first sight. There is limited help to look at streaming data in Production. You need to make it combine KPL, KCL and kinesis correctly. Also, we had to work much on increasing or reducing shards in real time. This is awesome service like many others you have but material and support is NOT that great. May be lately it has which I haven’t checked. Review collected by and hosted on G2.com.

The scalability and ease of use. The KCL and KPL libraries along with the AWS SDK make it really easy to work with, Review collected by and hosted on G2.com.

Working with shard iterators is not so great. Review collected by and hosted on G2.com.

We chose kinesis because of the ability to consume messages with the right sharding.and a managed serviced . Review collected by and hosted on G2.com.

1 megabyte limit, in our use case we have files that go beyond 1 mg even after compression. Review collected by and hosted on G2.com.

Easy to set up and throttle the size/messages limits Review collected by and hosted on G2.com.

The size limit can be further increased and the frequency can be reduced to less than a minute Review collected by and hosted on G2.com.