Re: Time to start publishing Spark Docker Images?

2021-08-16 Thread Holden Karau
These are some really good points all around. I think, in the interest of simplicity, well start with just the 3 current Dockerfiles in the Spark repo but for the next release (3.3) we should explore adding some more Dockerfiles/build options. On Mon, Aug 16, 2021 at 10:46 AM Maciej wrote: > I

Re: Time to start publishing Spark Docker Images?

2021-08-16 Thread Maciej
I have a few concerns regarding PySpark and SparkR images. First of all, how do we plan to handle interpreter versions? Ideally, we should provide images for all supported variants, but based on the preceding discussion and the proposed naming convention, I assume it is not going to happen. If

Spark docker image convention proposal for docker repository

2021-08-16 Thread Mich Talebzadeh
Hi, I propose that for Spark docker images we follow the following convention similar to flink as shown in the attached file So for Spark we will have -- 3.1.2-scala_2.12-java11 3.1.2_sparkpy-scala_2.12-java11 3.1.2_sparkR-scala_2.12-java11 If this makes