bluewave-labs / bluewave-labs/Checkmate

MongoDB 8.2.5 (tcmalloc-google) SIGSEGV crash loop on Linux kernel 6.19

Open
#3,444 0 comments 3 reactions 0 assignees View on GitHub
Dominant language
TypeScript
Stars
10.8k
Forks
1.2k
Avg merge
1d 4h
Merged PRs (30d)
74

Description

Environment:
- MongoDB: 8.2.5 (tcmalloc-google allocator)
- Docker: latest
- OS: Fedora 43
- Kernel tested: 6.19.x (broken), 6.18.7-200.fc43.x86_64 (stable)
- Filesystem: btrfs

Symptoms:
- MongoDB crashes with SIGSEGV (exit code 139, signal 11) every ~31 seconds
- Persistent crash loop: 1,705+ crashes over ~10 days
- Each crash triggers WiredTiger journal recovery on restart
- Zero crashes observed after downgrading to kernel 6.18.7

Impact:
- WiredTiger journal recovery on each restart generates btrfs CoW extents
- 2,000+ recovery cycles caused ~60GB ghost storage accumulation over 10 days

Tested:
- THP set to `always` (no effect)
- MongoDB stable on kernel 6.18.7 with identical config/data

Bisect hint:
- Regression introduced between kernel 6.18.7 and 6.19.x

Contributor guide

Open the contributing guide

Research direction

No repository file or test is identified. Start by reproducing the MongoDB Docker crash on Linux kernel 6.19.x and comparing it with 6.18.7 using the stated Fedora 43, btrfs, and tcmalloc-google configuration; done requires isolating the regression and identifying an actionable fix or upstream report.

Written by the indexing model from the issue text.

Assessment

Tech stack
docker, linux, mongodb
Domain
databases, infrastructure, operating-systems
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.