OData compliant API for Spark

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

OData compliant API for Spark

Affan Syed
All,

We have been thinking about exposing our platform for analytics an OData server (for its ease of compliance with 3rd party BI tools like Tableau, etc) -- so Livy is not in the picture right now.  

Has there been any effort on this regards? Is there any interest or has there been any discussion that someone can point towards? 

We want to expose this connection over API so the JDBC->thriftserver->Spark route is not being considered right now. 


- Affan
jgp
Reply | Threaded
Open this post in threaded view
|

Re: OData compliant API for Spark

jgp
I was involved in a project like that and we decided to deploy the data in https://ckan.org/. We used Spark for the data pipeline and transformation. Hih. 

jg


On Dec 4, 2018, at 21:14, Affan Syed <[hidden email]> wrote:

All,

We have been thinking about exposing our platform for analytics an OData server (for its ease of compliance with 3rd party BI tools like Tableau, etc) -- so Livy is not in the picture right now.  

Has there been any effort on this regards? Is there any interest or has there been any discussion that someone can point towards? 

We want to expose this connection over API so the JDBC->thriftserver->Spark route is not being considered right now. 


- Affan
Reply | Threaded
Open this post in threaded view
|

Re: OData compliant API for Spark

Affan Syed
Jean,

thanks for the pointer, but my question remains around building an odata server that maps onto REST. This can be a non-trivial effort, and I was hoping for people to jump in some discussion and maybe we can decide on how to execute in open source rather than doing an insource effort. 

To reiterate, the core benefit will be for users exposing their Big Data storage to external BI tools --- adding value in the case of PaaS use cases. 





- Affan


On Wed, Dec 5, 2018 at 6:50 PM Jean Georges Perrin <[hidden email]> wrote:
I was involved in a project like that and we decided to deploy the data in https://ckan.org/. We used Spark for the data pipeline and transformation. Hih. 

jg


On Dec 4, 2018, at 21:14, Affan Syed <[hidden email]> wrote:

All,

We have been thinking about exposing our platform for analytics an OData server (for its ease of compliance with 3rd party BI tools like Tableau, etc) -- so Livy is not in the picture right now.  

Has there been any effort on this regards? Is there any interest or has there been any discussion that someone can point towards? 

We want to expose this connection over API so the JDBC->thriftserver->Spark route is not being considered right now. 


- Affan