Kafka-on-Pulsar: bring native Kafka protocol support to Apache Pulsar

Architecture wise Pulsar is a far better piece of engineering than Kafka but with Confluent's funding and aggressive marketing it looks like Kafka has cornered the market for better or worse.

Re: Kafka-on-Pulsar: bring native Kafka protocol support to Apache Pulsar

#6

Post by adamfeldman » Tue, Mar 24, 2020, 10:42 PM UTC

tiew9Vii wrote: post #5

Does anyone know what is happening with Pulsar since Splunk brought them out? It looks to of stagnated since the announcement was made. Architecture wise Pulsar is a far better piece of engineering than Kafka but with Confluent's funding and aggressive marketing it looks like Kafka has cornered the market for better or worse.

Splunk bought a company called Streamlio [1].

StreamNative is the company behind the linked blog post and project.

My (outsider) understanding is that both Streamlio and Splunk employ(ed) active members of the Pulsar community.

[1]: https://www.splunk.com/en_us/blog/leadership/splunk-to-expan...

Re: Kafka-on-Pulsar: bring native Kafka protocol support to Apache Pulsar

#7

Post by oweiler » Wed, Mar 25, 2020, 4:34 PM UTC

tiew9Vii wrote: post #5

Does anyone know what is happening with Pulsar since Splunk brought them out? It looks to of stagnated since the announcement was made. Architecture wise Pulsar is a far better piece of engineering than Kafka but with Confluent's funding and aggressive marketing it looks like Kafka has cornered the market for better or worse.

I always hear it's better architecture-wise but what are the real life implications of difference?

Also, Kafka has:

* ksql * ksqlDB * Kafka Connect * Kafka Streams

Does Pulsar has something comparable?

Re: Kafka-on-Pulsar: bring native Kafka protocol support to Apache Pulsar

#8

Post by thenaturalist » Thu, Mar 26, 2020, 3:41 PM UTC

oweiler wrote: post #7

tiew9Vii wrote: post #5

Does anyone know what is happening with Pulsar since Splunk brought them out? It looks to of stagnated since the announcement was made. Architecture wise Pulsar is a far better piece of engineering than Kafka but with Confluent's funding and aggressive marketing it looks like Kafka has cornered the market for better or worse.

I always hear it's better architecture-wise but what are the real life implications of difference? Also, Kafka has: * ksql * ksqlDB * Kafka Connect * Kafka Streams Does Pulsar has something comparable?

Yes, you can run SQL using Presto [1], and by connecting. Not sure about ksqlDB as I am not familiar with it, but my understanding is Apache Flink would give you analog capabilities. Pulsar has IO capabilities analog to Kafka Connect [2]. Analogous to Kafka Streams, Pulsar offers Functions [3].

[1] http://pulsar.apache.org/docs/en/sql-overview/

[2] http://pulsar.apache.org/docs/en/io-overview/

[3] http://pulsar.apache.org/docs/en/functions-overview/