Apache Flink streaming features summary

[Flavio Pompermaier]

Hi to all,

I saw this comparison vs Spark, Storm and Apache Apex.

Since I didn't have time to look at the streaming part of Apache Flink, it could be interesting to have such a table (at least for the streaming features) also for Flink documented somewhere..what do you think?

Best,

Flavio

[Fabian Hueske-2]

Can you share a link to the comparison?

Thanks, Fabian

2015-09-28 9:32 GMT+02:00 Flavio Pompermaier <[hidden email]>:

Hi to all,

I saw this comparison vs Spark, Storm and Apache Apex.

Since I didn't have time to look at the streaming part of Apache Flink, it could be interesting to have such a table (at least for the streaming features) also for Flink documented somewhere..what do you think?

Best,

Flavio

[Flavio Pompermaier]

sorry, I forgot to paste it :(

https://www.datatorrent.com/project-apex/

On Mon, Sep 28, 2015 at 10:10 AM, Fabian Hueske <[hidden email]> wrote:

Can you share a link to the comparison?

Thanks, Fabian

2015-09-28 9:32 GMT+02:00 Flavio Pompermaier <[hidden email]>:

Hi to all,

I saw this comparison vs Spark, Storm and Apache Apex.

Since I didn't have time to look at the streaming part of Apache Flink, it could be interesting to have such a table (at least for the streaming features) also for Flink documented somewhere..what do you think?

Best,

Flavio

[Stephan Ewen]

Hi Flavio!

These marketing-style comparisons are almost always either wrong, or heavily tainted to make one particular project look good. They are pretty bad style for Apache projects, in my opinion. For that reason, we try to make more technical statements and comparisons.

There are quite some presentations out there that compare different streaming systems technically (and performance).

This set of slides by Gyula and Marton has a good comparison:

http://www.slideshare.net/GyulaFra/largescale-stream-processing-in-the-hadoop-ecosystem

Several blog posts have good comparisons as well

  - http://data-artisans.com/high-throughput-low-latency-and-exactly-once-stream-processing-with-apache-flink/

  - http://data-artisans.com/batch-is-a-special-case-of-streaming/

In general, the set of features and capabilities in Flink's streaming API is pretty rich and needs not shy comparison.

Greetings,
Stephan

On Mon, Sep 28, 2015 at 10:20 AM, Flavio Pompermaier <[hidden email]> wrote:

sorry, I forgot to paste it :(

https://www.datatorrent.com/project-apex/

On Mon, Sep 28, 2015 at 10:10 AM, Fabian Hueske <[hidden email]> wrote:

Can you share a link to the comparison?

Thanks, Fabian

2015-09-28 9:32 GMT+02:00 Flavio Pompermaier <[hidden email]>:

Hi to all,

I saw this comparison vs Spark, Storm and Apache Apex.

Since I didn't have time to look at the streaming part of Apache Flink, it could be interesting to have such a table (at least for the streaming features) also for Flink documented somewhere..what do you think?

Best,

Flavio

... [show rest of quote]

[Sebastian Schelter-2]

> These marketing-style comparisons are almost always either wrong, or
> heavily tainted to make one particular project look good. They are
> pretty bad style for Apache projects, in my opinion. For that reason, we
> try to make more technical statements and comparisons.

+1

[Flavio Pompermaier]

Ok, thanks for the pointers Stephan!

Best,

Flavio

On Mon, Sep 28, 2015 at 4:56 PM, Sebastian <[hidden email]> wrote:

These marketing-style comparisons are almost always either wrong, or
heavily tainted to make one particular project look good. They are
pretty bad style for Apache projects, in my opinion. For that reason, we
try to make more technical statements and comparisons.

+1