[SUBMARINE-232] [Umbrella] Custom Zeppelin Submodule

• Type: New Feature
• Status: Resolved
• Priority: Major
• Resolution: Won't Fix
• Component/s: Submodule Zeppelin
• Target Version: 0.3.0
• Assignee: Liu Xun
• Reporter: Liu Xun
• Created: 11/Oct/19 11:16
• Updated: 16/Dec/19 07:45
• Resolved: 16/Dec/19 07:45

Description

2. Interpreter
Submarine provides multi-language execution for notebooks in workbench through interpreters in a variety of programming languages (for example, python interpreter, spark interpreter).
2.1 Interpreter execution flow
1. The user enters the code in the workbench notebook and clicks the execute button. First the workbench server starts the interpreter process.
2. The workbench server sends the user's code to the interpreter process via the interface for execution, and gets the execution result, returning to the workbench notebook for display.

2.2 Integrate Zeppelin Interpreter
Zeppelin has this mature multilingual interpreter that can be introduced for use as a Submarine interpreter, but Submarine has some specific requirements for the use of the interpreter, such as Submarine's need to extend the interface of the Zeppelin interpreter, etc., so Zeppelin is required. The interface layer of the interpreter is modified to some extent.
At the same time, in order to be consistent with the development of the Zeppelin interpreter, we used submodule to customize the Zeppelin interpreter to a certain extent.

Design Doc: https://docs.google.com/document/d/1LJc5hRTWfIs6K5f7vMgGC_FUizeQf-qqiWICqO8SoR8/edit#

Issue Links

• is related to: SUBMARINE-165 [Umbrella] Submarine Workbench
• is related to: SUBMARINE-230 [Umbrella] Submarine Interpreter Module

Sub-Tasks

• [SUBMODULE] Delete unused modules In hadoopsubmarine/zeppelin repo: Resolved