k.rennerandClaude Opus 4.7 976d8b4dca immich: HPA for machine-learning ...

Bursty CPU profile (library scans, facial recognition, smart search
queries) that idles between. Cache PVC is RWX, so multi-replica is
safe. Cap at 2 to limit aggregate model RAM. Slow scale rates and
long scaleDown stabilization (10m) avoid churning the model cache,
which takes ~30s of cold-start to rebuild.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

2026-05-14 20:36:22 +02:00

.vscode

Add Unifi and Cloudflare provider configurations and examples

2026-02-02 21:54:10 +01:00

.woodpecker

registry-mirror: decommission LXC 250

2026-05-14 19:54:05 +02:00

ansible

registry-mirror: decommission LXC 250

2026-05-14 19:54:05 +02:00

ci

revert(security-fix-agent): drop image-tag-bump role

2026-05-01 15:09:06 +02:00

dns-server

chore: Remove local markdown docs; wiki is single source of truth

2026-04-21 07:42:43 +02:00

kubernetes

immich: HPA for machine-learning

2026-05-14 20:36:22 +02:00

opentofu

registry-mirror: decommission LXC 250

2026-05-14 19:54:05 +02:00

opentofu-bootstrap

rustfs(bootstrap): stop the three RustFS LXCs

2026-05-09 19:14:59 +02:00

scripts

import-cluster-access: default AWS_PROFILE to "secrets"

2026-05-06 15:49:25 +02:00

.gitignore

ci(security-scan): auto-fix agent that opens PRs for Trivy findings

2026-05-01 09:28:59 +02:00

S

Description

No description provided

38 MiB

Releases 1

master

Latest

2026-01-31 15:05:31 +00:00

Languages

Go 35.3%

Shell 24.5%

Python 22.5%

Jinja 5.1%

HCL 4.8%

Other 7.7%