lablup / lablup/backend.ai

Add configuration to limit the max number of container processes

Open
#12,238 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
670
Forks
183
Avg merge
15h 13m
Merged PRs (30d)
368

Description

## Motivation

- See [https://lablup.atlassian.net/browse/BA-6508](https://lablup.atlassian.net/browse/BA-6508).

## Required Features

- Add the following **per-resource-group** configuration for super-admins:
- A flag to enable or disable the pid limits. (default: on)
- The maximum number of processes. (default: 4096)

## Impact

- Users can no longer spawn indefinite number of processes, which may degrade the system performance and affect the workload sharing the same host.

## Testing Scenarios

- Let’s add unit tests for agents that sets the pid limit as a small number and executes a script within a kernel container that spawns many child processes above the limit.

JIRA Issue: BA-6521

Contributor guide

Open the contributing guide

Research direction

Start by reading the agent unit tests and the code that launches kernel containers; the issue does not name specific files or entry points. Trace how per-resource-group settings reach agents, then test a small process limit with a script that spawns children. Done means super-admins can toggle the limit and set its maximum, with defaults of on and 4096, and the agent enforces it.

Written by the indexing model from the issue text.

Assessment

Tech stack
docker, python
Domain
backend, infrastructure
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
52/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.