March 18, 2016 at 6:35 am #13169
Can anyone comment on how Kamanja compares with CASK.io?
CASK seems to simplify application development on Hadoop and it can integrate with Spark etc.
Can Kamanja utilize CASK?
Is there any relationship between these two?
March 18, 2016 at 6:37 am #13172
Thanks for asking such an interesting question.
They are functionally quite different.
Cask.io provides an abstraction layer that sits on top of multiple storage implementations, Hbase, Hadoop, Hdfs and so on. It allows people developing applications to use the Cask.io abstraction layer to develop applications. Because of the they applications will be portable to a larger number of use cases. Hadoop itself can look quite different depending on how it is configured so just because two sites are running Hadoop is no guarantee they are the same (or even close).
Kamanja manages an arbitrarily large amount of input that is coming at it in real-time. For instance, bank transactions, geographic location, exercise information and countless other streams of information that have become part of our day-to-day life. Kamanja identifies and separates the individual streams and analyzes them in some way. “Is a bank balance too low?” “Has a person change location?” “Has the person exercised more than the past?”
These are simple, familiar examples. Creating the analysis in the past was time-consuming. It would take weeks or even months. Kamanja allows that process to be shortened to days. The piece that does the analysis is called a “model.” With Kamanja you can pop out a model and plug in a new one and the rest of the system does not need to be changed. The system can grow by adding more nodes to handle more data. Changing what it is analyzing is just a matter of changing the models.
There are ways that we can interact with Cask.io. Since we are both open-source it seems like an opportunity for someone to explore the possibilities. We are starting a Kamanja User’s Group (watch here for times). It will be a great place to meet with others and explore ways that the rapidly growing open-source ecosystem can work together in the exciting areas that are evolving in the big-data world.