bottlerocket-os/bottlerocket

Consider adding `systemd-coredump` support?

Open

#2,858 opened on Mar 6, 2023

View on GitHub
 (2 comments) (6 reactions) (0 assignees)Rust (6,938 stars) (386 forks)batch import
area/corehelp wantedstatus/needs-proposaltype/enhancement

Description

What I'd like:

A way to have some access/management of core dumps, but without disks filling up, or needing to do hacky things to work around container filesystem resets deleting container images.

It looks like there is a reasonable solution implemented via systemd's coredump process that allows storing and fetching core dumps in a centeralized way, as well as being able to do things like compression and disk usage management. Obviously this can impact a lot of things in the system just from resource usage, but it seems like an option for a coordinated approach to core dump management.

My main environment of concern is Kubernetes, but this likely would apply to ECS as well.

Any alternatives you've considered:

Status-quo: No core dumps preserved. Allow programs dumping core to disk till disk runs out. Hack something in with a bootstrap container.

Edit: Submitted too quickly.

Contributor guide

Tech stack
rust
Domain
backendinfrastructure
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
over 1 week
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
clear
Prerequisites
Understanding of systemd and coredumpContainer filesystem conceptsFamiliarity with Bottlerocket OSRust programming experience
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
25
Research direction
Review systemd coredump configuration and API for storing and fetching core dumps. Investigate Bottlerocket's current core dump handling (likely via kernel.core pattern). Explore integration points in the OS, such as adding a systemd service or modifying existing ones. Consider disk usage management, compression, and container filesystem resets. Look into existing solutions in other container optimized OSes. The issue discusses Kubernetes and ECS environments, so research how these orchestration platforms interact with node level core dumps. Check for any prior work like PRs or branches in the Bottlerocket repository. The 2 comments may provide additional context; read them for further direction.