Skip to main content
Back to Journal
user@argobox:~/journal/2026-01-29-81mb-of-markdown
$ cat entry.md

81MB of Markdown

○ NOT REVIEWED

81MB of Markdown

Date: January 29, 2026 Duration: ~20 minutes once I found it Problem: Obsidian goes black and crashes during indexing


The Symptom

Open Obsidian. Watch it load the vault. Ten seconds later... black screen. Crash.

Every. Single. Time.

This was my argobox-technical-vault. 181MB of documentation, session notes, project files. Important stuff.


The Cause

Two files:

File Size
chatgpt-homelab-conversations-2023-2024.md 49MB
claude-homelab-conversations-2025.md 32MB

81MB of markdown that Obsidian was trying to index.

I'd been dumping conversation archives into the vault. "For reference." Because someday I might search them.

Obsidian's indexer disagreed with this plan.


The Fix

Moved them out:

mkdir -p $HOME/Vaults/archive-test/

mv Vaults/argobox-technical/sessions/chatgpt-homelab-conversations-2023-2024.md \
   Vaults/archive-test/

mv Vaults/argobox-technical/sessions/claude-homelab-conversations-2025.md \
   Vaults/archive-test/

Added them to .obsidianignore just in case:

# Skip massive conversation archives
sessions/chatgpt-homelab-conversations-2023-2024.md
sessions/claude-homelab-conversations-2025.md

Vault size dropped from 181MB to 101MB.

Obsidian opens fine now.


The Lesson

Obsidian is not a conversation archive. It's a knowledge base.

If I need to search old conversations, that's what grep is for.


Also Fixed Today

While I was in infrastructure mode, I noticed an old remote-access DNS record still pointed at a legacy public endpoint.

Changed RustDesk to use the private overlay path instead:

services:
  hbbs:
    command: hbbs -r overlay-peer.internal:21117

The cleanup item is to remove the stale external DNS record after confirming all clients use the overlay route.