Apache Kafka is an open-source stream processing platform developed by the Apache Software Foundation written in Scala and Java. The Kafka event streaming platform is used by thousands of companies for high-performance data pipelines, streaming analytics, data integration, and mission-critical applications.
Apache Kafka is built for scale. From high throughput and real-time data streaming, it has a strong advantage over RabbitMQ with its low latency. This put Apache Kafka at the forefront as the platform of choice for large datasets messaging and ensuring scalability when data …
Verified User
Employee
Chose Apache Kafka
It had the clustering functionality and gave tolerance against machine failure.
Apache Kafka can work at a higher scale as compared to SQS. It can work with higher size per message and millions of messages per second. Moreover it can be scaled horizontally by adding more brokers to the cluster. SQS is good enough for simple use cases like making a task …
Verified User
Consultant
Chose Apache Kafka
I used other messaging/queue solutions that are a lot more basic than Confluent Kafka, as well as another solution that is no longer in the market called Xively, which was bought and "buried" by Google. In comparison, these solutions offer way fewer functionalities and respond …
Apache Kafka is open-sourced, scales great has cloud agnostics and performs better than Amazon Kinesis [in my view]. Amazon Kinesis has some limitations and vendor lockin is not something I [like]. With Confluent operators you can easily install it on a kubernetes cluster.
For us, Kafka really doesn't have a 1:1 alternative. We have used ActiveMQ extensively and we still use it as a lighter option for small messages. The situation is similar with Redis - although it could be used like a Kafka alternative, we do use it just as a per-component …
We really needed to get away from using a SQL database to act as a queue for processing records, so a new solution was needed. Kafka is a leading software application initially designed for queuing messages which is essentially what we were looking for. It has a great user …
Apache Kafka is much more scalable and more reliable. Does not depend on memory, works well on rotational disks and that makes it a cheaper to use solution on low hardware requirements. Running multiple consumers on the same topic can also mean processing the same data again …
All stack tech helps our app and system. These technologies allow us to have the data available faster between different regions (due to our particular configuration) and thus the data and processing load of each system is lower. This allows the systems to be used more …
Kafka is not a real messaging broker implementation as RabbitMQ or TIBCO EMS/JMS are. Although it can be used as messaging, we like the idea behind the Kafka (data isn't "passing by," instead it remains centra, so the client can revisit the data if necessary). This also …
Confluent Cloud is still based on Apache Kafka but it has a subscription fee so, from a long term perspective, it is wiser to deploy your own Kafka instance that spans public and private cloud. Amazon Kinesis, Google Cloud Pub/Sub do not do well for a very number of messages …
I would only use RabbitMQ over Kafka when you need to have delay queues or tons of small topics/queues around. I don't know too much about Pulsar - currently evaluating it - but it's supposed to have the same or better throughput while allowing for tons of queues. Stay tuned - I …
Kafka is faster and more scalable, also "free" as opensource (albeit we deploy using a commercial distribution). Infrastructure tends to be cheaper. On the other hand, projects must adapt to Kafka APIs that sometimes change and BAU increases until a major 1.x version comes out …
Apache Kafka is well-suited for most data-streaming use cases. Amazon Kinesis and Azure EventHubs, unless you have a specific use case where using those cloud PaAS for your data lakes, once set up well, Apache Kafka will take care of everything else in the background. Azure EventHubs, is good for cross-cloud use cases, and Amazon Kinesis - I have no real-world experience. But I believe it is the same.
Really easy to configure. I've used other message brokers such as RabbitMQ and compared to them, Kafka's configurations are very easy to understand and tweak.
Very scalable: easily configured to run on multiple nodes allowing for ease of parallelism (assuming your queues/topics don't have to be consumed in the exact same order the messages were delivered)
Not exactly a feature, but I trust Kafka will be around for at least another decade because active development has continued to be strong and there's a lot of financial backing from Confluent and LinkedIn, and probably many other companies who are using it (which, anecdotally, is many).
Sometimes it becomes difficult to monitor our Kafka deployments. We've been able to overcome it largely using AWS MSK, a managed service for Apache Kafka, but a separate monitoring dashboard would have been great.
Simplify the process for local deployment of Kafka and provide a user interface to get visibility into the different topics and the messages being processed.
Learning curve around creation of broker and topics could be simplified
Apache Kafka is highly recommended to develop loosely coupled, real-time processing applications. Also, Apache Kafka provides property based configuration. Producer, Consumer and broker contain their own separate property file
Support for Apache Kafka (if willing to pay) is available from Confluent that includes the same time that created Kafka at Linkedin so they know this software in and out. Moreover, Apache Kafka is well known and best practices documents and deployment scenarios are easily available for download. For example, from eBay, Linkedin, Uber, and NYTimes.
I used other messaging/queue solutions that are a lot more basic than Confluent Kafka, as well as another solution that is no longer in the market called Xively, which was bought and "buried" by Google. In comparison, these solutions offer way fewer functionalities and respond to other needs.
Positive: Get a quick and reliable pub/sub model implemented - data across components flows easily.
Positive: it's scalable so we can develop small and scale for real-world scenarios
Negative: it's easy to get into a confusing situation if you are not experienced yet or something strange has happened (rare, but it does). Troubleshooting such situations can take time and effort.