pytorch / pytorch/tutorials

💡 [REQUEST] - Declare warning in tutorial that following distributed tutorial may expose tcp ports.

Open
#3,323 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

distributed
Dominant language
Python
Stars
9.3k
Forks
4.4k
Avg merge
1d 21h
Merged PRs (30d)
4

Description

🚀 Describe the improvement or the new tutorial

Add a warning or declare that setting of the environment variable "NCCL_SOCKET_IFNAME" is required to prevent opening up sockets outside of your local machine even when setting the master address to "localhost"

Existing tutorials on this topic

https://pytorch.org/tutorials/intermediate/FSDP_tutorial.html
https://pytorch.org/tutorials/intermediate/ddp_tutorial.html

Additional context

No response

cc @wconstab @osalpekar @H-Huang @kwen2501

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Review the FSDP tutorial and DDP tutorial linked in the issue, focusing on the distributed setup that uses the NCCL_SOCKET_IFNAME environment variable. Add a warning explaining the socket and TCP-port exposure risk and the stated setting needed to prevent it, then verify both tutorial pages contain the warning.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
distributed-systems, documentation, security
Issue type
Documentation
Difficulty
2/5
Estimated time
Half a day
Activity status
Stale
Clarity
Clearly specified
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.