python-debugpy

Remote Python debugging combines pdb REPL for local inspection with debugpy DAP to attach to running processes. You configure debugpy to listen on a port, attach your client, set breakpoints, and control execution across local scripts and daemons.

Yes, you can attach a debugger to a running Python process for post-mortem analysis using debugpy. This allows you to inspect state after a failure, analyze long-running services, and diagnose test failures without restarting the application.

The best way to debug long-running Python daemons is using debugpy remote attachment via DAP. You inject debugpy into the running process, connect a DAP client, set breakpoints, and inspect variables without interrupting the daemon's startup cycle.

No, you do not need debugpy installed to use pdb for local breakpoints. pdb is built into Python, but debugpy is required separately when you need to attach to running processes or perform remote DAP debugging.

To start a Python script under pdb, you launch the application with pdb enabled or insert breakpoint() calls in your code. This pauses execution and opens an interactive REPL for local variable inspection and step-by-step control.

Debugpy DAP is used instead of pdb when you need remote attachment to running processes or cross-process analysis. pdb limits you to local terminal REPL, while debugpy supports connecting external clients for complex scenarios like long-running services.