Apache Kafka Review

Interactions among micro-services are used as input to our analytics infrastructure.

How has it helped my organization?

Kafka was at the base of our system architecture. The system was designed as an event based architecture. Almost all the interactions among micro-services and the same data are used as input to our analytics infrastructure.

What is most valuable?

  • Scalability
  • Reliability
  • Ease of use

What needs improvement?

Stability of the API and the technical support could be improved.

The Kafka API is changing quite radically with the different releases. There are many new improvements and that's good. But the inherent cost of adapting to a new version of the platform was worrying me at the time.

The documentation was sometimes misleading, since it was describing some feature in the new version of the API rather than the one we were using.

What do I think about the stability of the solution?

We did not encounter any issues with stability.

What do I think about the scalability of the solution?

We did not encounter any issues with scalability.

How is customer service and technical support?

We were not completely satisfied with the technical support. We subscribed to the Confluent professional platform to receive guidance and support on development and deployment. Whilst the development side is quite well covered by their consultants, the deployment and administration is not at the same level.

Which solutions did we use previously?

The previous solution was not really an equivalent one. I have been using several messaging systems, but Kafka fits us better for a more scalable system.

How was the initial setup?

The initial setup was straightforward.

What's my experience with pricing, setup cost, and licensing?

I would not subscribe to the Confluent platform, but rather stay on the free open source version. The extra cost wasn't justified.

Which other solutions did I evaluate?

We didn't evaluate other options, as we already had a positive experience across the team with Kafka. Everybody agreed to work with it.

We were considering Kinesis too, since we were running on AWS. We preferred to opt for a tool with which people were more familiar.

What other advice do I have?

The product is easy to use. However, to leverage its power, there is a need for good knowledge of event based processing. I suggest using the massive amount of material shared by the Confluent team, or what is available online.

Disclosure: I am a real user, and this review is based on my own experience and opinions.
Add a Comment
Sign Up with Email