Apache Kafka Reviews

We are building solutions on Apache Kafka for four customers. The customers we have are in various sectors, such as healthcare and architecture.

Apache Kafka has good integration capabilities and has plenty of adapters in its ecosystem if you want to build something. There are adapters for many platforms, such as Java, Azure, and Microsoft's ecosystem. Other solutions, such as Pulsar have fewer adapters available.

I have been using Apache Kafka for three years.

Apache Kafka is stable.

I would recommend Apache Kafka for any enterprise.

The amount of people using the solution depends on the application. However, the starting point is from 6,000 to 7,000 concurrent users.

There is not any support, Apache Kafka is open-source.

We have been experimenting with other solutions such as VMware RabbitMQ and Pulsar.

We are going to replace the Apache Kafka solution using Pulsar.

Pulsar gives more scalability to an even grouping, but Apache Kafka is used more if you want to send something in a time series-based. If this does not matter to you then Pulsar could be more customizable. Apache Kafka is nothing but a streaming system with local storage. Apache Kafka fits into many use cases, it's very direct, but if you want more specific use cases and you use Apache Kafka, Pulsar could be considered.

Apache Kafka was simple to install. If you have a complicated clustered production, it takes time. However, for the development, it doesn't take more than one or two hours.

We have approximately two to four technical managers that are deploying and supporting Apache Kafka. A technical manager is necessary.

Apache Kafka is an open-sourced solution. There are fees if you want the support, and I would recommend it for enterprises. There are annual subscriptions available.

Apache Kafka is one of the best open-source solutions that are available today.

I would recommend this solution to others.

I rate Apache Kafka an eight out of ten.

Kafka has a guaranteed delivery mechanism that is very easy to set up. When starting out with minimal hardware, it can handle very large data volumes. When prototyping and creating a proof of concept, Kafka has helped to speed up the timeline from the prototype all the way to production volumes.

Apache Kafka is actually a distributed commit log. That is different than most messaging and queuing systems before it. I find the ability to write data at one velocity and have subscribing consumers read at different velocities to be the best feature.

The GUI tools for monitoring and support are still very basic and not very rich. There is no help in determining a shard key for performance.

We did not have any issues with stability.

We did not have any issues with scalability.

• Kafka is open source from LinkedIn and support comes from the community of users.
• You can go with Confluent, the company that was founded by the original engineers from LinkedIn.
• You can go with a cloud hosting service, like AWS EMR or Azure HDInsight.

We used traditional message queues and file semaphores. There was a lot of overhead with asynchronous messages being put into an order and making sure nothing got dropped. It required a lot of code and maintenance.

Since it is open source, you are on your own for setup. However, the tutorials from the Apache foundation and online sources have been an immense help.

Getting started is very easy. The complexity of very large volumes of data and appropriate sharding, however, is difficult. There are fewer resources for tuning and best practices.

When starting to look at a distributed message system, look for a cloud solution first. It is an easier entry point than an on-premises hardware solution. A lot of the complexity has already been taken care of. Both AWS and Azure have supported Kafka clusters that can be provisioned very easily.

We looked at RabbitMQ and Spark Streaming.

Be sure to define the use cases as best as possible at first.

Kafka is very good, but it is complex to support. It can handle any message size, whereas native cloud options have size limitations.

Be sure to understand what messages will be sent and how many discrete topics will be needed.

Be aware that you must code both producers and consumers.

The bulk of the work is with the consumer.

The Apache stack for Kafka is very open source. There are essentially no tools other than command line options to monitor brokers and topic health. So there are 3rd party tools that will help with that, some free, some paid – but it requires that you install agents on the servers hosting Kafka and open up ports for netbeans on the scripts that start up the Kafka services. Additionally, you also have to monitor zookeeper – which is very memory intensive. Cloud offerings that provide the whole modern data architecture stack – like AWS EMR and Azure HDInsight as well as Hortonworks and Cloudera provide a console GUI as part of each of their offerings. Also Confluent, a company founded by the Linked-In engineers that designed Kafka, also have a paid enterprise offering that has much better tools for maintain the kafka cluster. But apache Kafka with the community – you are on your own.

I use Kafka to send network packets from different sources to my cluster. We have around 10 users at my company.

Kafka is scalable. It can manage a high volume of data from many sources.

The interface has room for improvement, and there is a steep learning curve for Hadoop integration. It was a struggle learning to send from Hadoop to Kafka. In future releases, I'd like to see improvements in ETL functionality and Hadoop integration. 

I have used Kafka for around six months.

I rate Apache Kafka seven out of 10 for stability. 

I rate Kafka eight out of 10 for scalability. 

I rate Apache support six out of 10. It was hard to find the information I needed. 

Neutral

Before Kafka, I sent feeds directly to Hadoop.

I initially found Kafka difficult to set up, so I would rate it about five out of 10 for ease of setup. After I learned more about the platform, I would rate it eight out of 10. It is deployed on-premises over a cluster of three or four PCs. You can deploy Kafka in a few hours with one person. 

Kafka is open source. 

I rate Apache Kafka eight out of 10. I would recommend it to others. 

Our primary use case for this solution is streaming.

We have been using this solution for four years.

The solution is stable. However, it's not possible to substitute IBM MQ with Apache Kafka because the JMS part is not very stable. It is inadequate and doesn't have the support of the MQI interface of IBM MQ.

The solution is scalable. Deployment is speedy, but we don't have many installations. We have over a thousand users using this solution and will most likely increase the number of users because we have tested 100,000 messages per second. The solution is impressive.

We previously used Mosquitto and Rabbit solutions, but we currently use Apache Kafka.

We are licensed annually for this solution.

I rate this solution a nine out of ten for streaming. I recommend it to other people. The solution is good, but its performance can be improved.

It's a combination of an on-premise and cloud deployment. We use AWS, and we have our offshore deployment that's on-premise for OpenShift, Red Hat, and Kafka. Red Hat provides managed services and everything. We use Kafka and a specific deployment where we deploy on our basic VMs and consume Kafka as well.

We publish or stream all our business events as well as some of the technical events. You stream it out to Kafka, and multiple consumers develop a different set of solutions. It could be reporting, analytics, or even some data persistence. Later, we used it to build a data lake solution. They all would be consuming the data or events we are streaming into Kafka.

With Kafka, events and streaming are persistent, and multiple subscribers can consume the data. This is an advantage of Kafka compared to simple queue-based solutions.

We are still on the production aspect, with our service provider or hyper-scalers providing the solutions. I would like to see some improvement on the HA and DR solutions, where everything is happening in real-time. 

Kafka's interface could also use some work. Some of our products are in C, and we don't have any libraries to use with C. From an interface perspective, we had a library from the readies. And we are streaming some of the products we built to readies. That is one of the requirements. It would be good to have those libraries available in a future release for our C++ clients or public libraries, so we can include them in our product and build on that.

We've been using Apache Kafka for the past two to three years.

Kafka is stable. It's a great product. 

We did some benchmarking, but we are still looking further to scale up some of the benchmarking and performances. So far, it meets all our business requirements. We are just developers, so everything goes to the clients, who will deploy it at their scale and use it for their end customers. So were are looking at it from a developer's perspective. Those who are developing the products are working on this.

We haven't really contacted technical support, but some of our clients have subscribed to support from the vendors. We generally look for open-source solutions. From there, we try to figure out if there are any issues. There's a good online community where you can ask questions.

We were able to deploy and use it with no problems for our use case. We didn't find it so complex. We work with so many applications, databases, Postgres, and so many other things, so we could manage it easily. We deployed Kafka in a few hours. We have an infrastructure team and DevOps. Those teams are pretty capable, and they've completely automated the whole deployment. It always takes time the first time you upgrade any application, not just Kafka. We might discover some issues, such as configuration, parameters, compatibility, etc. Once that becomes standard, it is stable, and then they only need to replicate it to the different environments or different developers groups. We have a sophisticated process.

I rate Apache Kafka eight out of 10. There are so many products on the market, so my advice is to consider if Kafka suits your business requirements first. If it's suitable, the next step is to check whether all the technical requirements are met. If everything checks out, I would say that Kafka is a relatively stable, sound, and scalable product, so they can try it out. 

We primarily use the solution for big data. We often get a million messages per second, and with such a high output we use Kafka to help us handle it. 

When we're working with big data, we need a throughput computing panel, which is something that Kafka provides, and something we find extremely valuable. It helps us support computing and ensures there's no loss of data. It can even do replication with some data.

The delivery of data is it's most valuable aspect.

It's an easy to use product overall.

The solution is quite mature.

It's an open-source product, which means it doesn't cost us anything to use it.

We're still going through the solution. Right now, I can't suggest any features that might be missing. I don't see where there can be an improvement in that regard.

The speed isn't as fast as RabbitMQ, even though the solution touts itself as very quick. It could be faster. They should work to make it at least as fast as RabbitMQ.

The UI is based on command line. It would be helpful if they could come up with a simpler user interface.

They should make it easier to configure items on the solution.

The solution would benefit from the addition of better monitoring tools.

I've been using the solution for six months.

The solution is a bit slow in comparison to RabbitMQ. It's supposed to be a very fast solution, and it has okay performance, but speed-wise, it's quite slow.

The scaling of the solution is quite good.

In terms of technical support, we don't get that directly from Apache Kafka. We have certain cloud data distribution so we get assistance from our cloud data support.

We're continuously deploying the product. We're still in the process of deployment.

It's an open-source product, so the pricing isn't an issue. It's free to use. We don't have costs associated with it.

I'm not the product owner, so I didn't have a say in what should be chosen. We were seeing a high throughput with Kafka which is why we ultimately chose it.

 I'd rate the solution eight out of ten. It's good at scaling, and, performance-wise, it's excellent. If they could add upon the UI and allow for easier configuration, I'd rate them higher.

It's convenient and flexible for almost all kinds of data producers. We integrated it with Kafka Streams, which can perform some easy data processing, like summary, count, group, etc

It eases our current data flow and framework, which digests all types of sources regardless of it being structured or not.

• High availability
• High throughput

With such a large digest, I was genuinely impressed at the process being almost real-time.

Kafka 2.0 has been released for over a month, and I wanted to try out the new features. However, the configuration is a little bit complicated: Kafka Broker, Kafka Manager, ZooKeeper Servers, etc.

The most valuable features are performance, persistent messaging, and reliability. It allows us to persist the message for a configurable number of days, even after it has been delivered to the consumer. The message delivery is also fast.

We wanted to track the customer activities on our application and store those details on another system(RDBMS/Apache Hadoop). We do extensive analysis with that. This helps the company to analyze the customer activities, such as search terms, and do better.

It’s perfect for our requirements.

I have been using Apache Kafka for two years.

We have had no issues with stability.

We have had no issues with scalability.

We use the open source one, so we did not opt for any technical support.

We started to use Apache Kafka with our application from scratch.

The initial setup was straightforward. We faced some issues during the development in areas such as message producer and consumer. We rectified those with the tweaking the producer and consumer configurations. The documentation is very good.

I don’t have any idea, as we use the open source version.

It's a high-performance distributed system. If you want to track the user activities or any stream processing, then this is perfect. We have used Docker Kafka for our implementation. It's very easy for setup and testing. You could also try the same.