Spark SQL FAQs

Question 1

What is Spark SQL and what is its primary function?

Accepted Answer

Spark SQL is an MCP (Model Context Protocol) server designed to empower AI assistants to query Spark SQL and other HiveServer2-compatible clusters (like AWS EMR, Hive, Impala, Presto) using the Thrift protocol. It acts as a secure intermediary for AI-driven data interactions.

Question 2

Can AI assistants discover database schemas using Spark SQL?

Accepted Answer

Yes, Spark SQL provides tools for schema discovery, enabling AI assistants to list available databases, list tables within a database, and describe the structure (columns, types) of specific tables.

Question 3

What security features does Spark SQL offer?

Accepted Answer

It enforces read-only SQL queries (only SELECT, SHOW, DESCRIBE, EXPLAIN, WITH), sanitizes database error messages to prevent data leakage, and handles credentials securely by masking passwords in logs and configurations. It also supports various secure authentication methods.

Question 4

Which authentication methods are supported by Spark SQL?

Accepted Answer

Spark SQL supports multiple authentication methods including NONE (for development), LDAP, Kerberos, CUSTOM, and NOSASL, allowing flexible integration with different enterprise security infrastructures.

Question 5

How do I integrate Spark SQL with my AI assistant environment (e.g., Claude Code)?

Accepted Answer

After installing via `pip`, you set environment variables for your Spark cluster host, port, and authentication. Then, you add the `spark-sql-mcp-server` command and its environment variables to your Claude Code or Claude Desktop MCP configuration file (`.claude.json` or `.claude/mcp.json`).

Spark SQL

Spark SQL

Key Features

Use Cases

Key Features

Use Cases