Does Apache Spark require droplet-to-droplet networking (Private networking)? and what is private networking on DigitalOcean?

February 2, 2018 1.3k views
Apache Big Data Clustering Ubuntu

Some background: First time building on the DigitalOcean droplet platform. Tried setting up a spark cluster with multiple droplets.

Issues that I ran into: The master could connect to the worker and in the workers log it said it connected to the master after 59ms. But then it runs into an issue where the worker cant hold the connection. The first time I did it was with private networking as I assumed having them all in the same region would be beneficial.

So im wondering if using private networking caused the issue? along with the questions in the title to clarify if I'm understanding the descriptions provided when creating a droplet.

Thank you

1 Answer

Dear User,

I will be able to help you in your issue could you please share the logs and also if it's possible to share the steps you performed to setup your cluster so that I can replicate it in my account?

Regards,
Sandeep Kumar

Have another answer? Share your knowledge.