Skip to content

Jupyter-Notebook-Setup.md

Latest commit

 

History

History
105 lines (74 loc) · 4.25 KB

Jupyter-Notebook-Setup.md

File metadata and controls

105 lines (74 loc) · 4.25 KB
Raw

Jupyter Notebook Setup

Installing and running Jupyter Notebook / JupyterLab inside a virtual environment for interactive security data analysis and PoC exploration.

Overview

Jupyter provides an interactive, cell-based environment ideal for exploratory work: parsing logs, triaging captured traffic, prototyping an exploit chain, or charting scan results. Code runs cell-by-cell with output (including tables and plots) shown inline, so you iterate fast without re-running a whole script. For security work, install Jupyter into a per-project [[Managing-Virtual-Environments|venv]] and register that venv as a kernel so notebooks use your pinned tooling.

Syntax

# Install into an activated venv
python3 -m venv .venv && source .venv/bin/activate
pip install jupyterlab

# Launch
jupyter lab            # full IDE-like interface
# or
jupyter notebook       # classic single-notebook UI

Explanation

Key ideas:

  • Kernel — the interpreter a notebook runs against. Register your venv as a named kernel so the notebook uses it rather than the global Python.
  • Cells — code or Markdown blocks executed independently; state persists in the kernel between cells.
  • JupyterLab vs Notebook — Lab is the modern multi-panel UI; classic Notebook is lighter.
# Register the active venv as a Jupyter kernel
pip install ipykernel
python -m ipykernel install --user --name acme-engagement \
    --display-name "Python (acme)"

# List and remove kernels
jupyter kernelspec list
jupyter kernelspec uninstall acme-engagement

Warning

jupyter lab starts a local web server. Do not bind it to 0.0.0.0 on an untrusted network — a Jupyter server with a leaked token grants remote code execution. Keep it on localhost and use SSH port-forwarding to reach a remote instance.

Examples

# In a notebook cell: parse an auth log and count failed SSH logins
import re
from collections import Counter

with open("/var/log/auth.log") as fh:
    ips = re.findall(r"Failed password.*from (\d+\.\d+\.\d+\.\d+)", fh.read())

Counter(ips).most_common(5)

Output

[('192.168.1.50', 214),
 ('10.0.0.9', 88),
 ('172.16.4.2', 31)]

Security Use Cases

  • Triaging large log or PCAP-derived datasets interactively, keeping intermediate results in memory across cells.
  • Prototyping an exploit or crypto attack step-by-step, inspecting each transformation before committing it to a script.
  • Producing a shareable, reproducible analysis notebook as a client deliverable, with narrative Markdown alongside the code.

Best Practices

  • Always install Jupyter into a project venv and register it as a named kernel; never sudo pip install jupyter.
  • Reach remote notebooks via ssh -L 8888:localhost:8888 user@host, never by exposing the port.
  • Clear sensitive output (tokens, dumped hashes) before saving/sharing a notebook — output is stored in the .ipynb.
  • Graduate stable notebook code into a proper module once it works (see [[Project-Structure-Best-Practices]]).

Common Mistakes

  • Running the notebook against the global kernel, so pinned venv packages are missing or wrong versions.
  • Binding the server to all interfaces or disabling token auth for convenience — a critical RCE exposure.
  • Relying on hidden cell-execution order; re-run "Restart & Run All" to confirm the notebook is reproducible.

Practical Lab

Goal: run a venv-backed notebook and analyze data interactively.

  1. Create and activate a venv, then pip install jupyterlab pandas.
  2. Register the kernel: python -m ipykernel install --user --name lab-kernel.
  3. jupyter lab, create a notebook, and select the lab-kernel.
  4. In cells, load a CSV of scan results with pandas and compute the count of open ports per host.

References

Related

  • [[Selecting-an-IDE]]
  • [[Managing-Virtual-Environments]]
  • [[Project-Structure-Best-Practices]]
  • [[Python-Environment-Setup/Readme|Python Environment Setup]] — module index
  • [[Readme|Python for Security Professionals]] — course home