We use it as part of a streaming pipeline. It helps with back-pressure when there is a sudden burst in the events and it also allows us to consume data in parallel.

easy integration with AWS lambda, kinesis firehose (s3) and third party software ( ex: splunk)

No autoscaling for now. For now we are setting alarms when there is an increase in the incoming data and then we scale it manually. We are thinking of automating this process with cloudwatch+ sns+ lambda

When it comes to availability, durability, and scalability, it is one of the best performant products for data streaming I have already used. It can also integrate with a lot of AWS services.

Other than the expertise in streaming data it needs, it also can incur high costs depending on the data stored and processed.

A well built advanced analytics tool with good scalability, maintainance and integration capabilities to make data more meaningful.

Used it as a basis for distributing information for IOT, multiple databases including SQL and NoSql, websites, that was used as a message queuing flow all put together. Can monitor all the IOT sensors information with its reporting and processing information. The data streaming events are fast and you can perform create, del, update and read (CRUD) operations on them, with good data retention. For the case of ML you can use it to find data anomalies.

Not much of a community base for doubt clearing, troubleshooting issues and solving problems. When went through the documentation, found it difficult to understand and confusing.

stream the data without any error or miss any dataeasy to setup the stream from the server

there is a delay of a stream of logs when integrate with another tool it is taken at least 5 min to sync each slot.

One of the best features of Amazon Kinesis is the easy integration of it with other AWS services. We often require the streaming data from Kinesis to be stored and later analyzed. This requires easy integration with services like Lambda, S3, Data lakes etc. This can be done very easily in the case of Kinesis.

There is very less visibility when the data is within AWS Kinesis. The user of the data has not idea where it resides. This is specially important in situations where there are drop in messages. Although its not an exact science Kinesis can make it better.

Amazon Kinesis is very powerful . You can gather data quickly in a very effective way

Like lot of AWS service, it can be a bit tricky to setup and documentation can be a bit complex