Local & private AI for security

Setups focused on locking down local AI services: sandboxing, resource limits and keeping exposed ports under control.

1 real example, each linked to its original source.

Sponsors
Rebel Bites+Your tool here
Advertise here

Hardening a two-GPU Ollama server with systemd: 9.2 UNSAFE to 1.4

A homelab write-up that adds measured memory and CPU limits, restart policy and sandboxing to Ollama through systemd drop-ins on a server with two RTX 5060 Ti cards, testing after every round that the GPUs still work.

OllamasystemdLlamaQwen#homelab#security
Gancho Stoychev (gshell.dev) · 2026-10-15