12 — Optimization Opportunities

Observations from the running infrastructure. Each item is independent — accept, defer, or reject.

For Nexa directly

Reuse, don't redeploy. The earlier DEPLOYMENT.md would have spun a second Memos / n8n / Qdrant. The current homelab already runs all three. The new 09-deployment treats these as pre-existing — keeps the config minimal and avoids port collisions.
Use LiteLLM virtual keys per logical caller. Today there's one SAIA key. Issuing one key per workflow (nexa-router, nexa-embed, nexa-digest) lets you set different per-key rate/cost limits and disable a single workflow without rotating everything.
Use n8n's credential objects, never inline secrets. The current workflows under nexa-core/n8n-workflows/phase-1/*.json should be reviewed — if any header Authorization is hardcoded, replace with credential references before importing.
Centralise system alerts on a single ntfy topic (nexa.system). Backrest, Proxmox notifications, n8n failure-webhook and the Octoprint Exited state all go to that topic; one Memos system memo aggregates them.
Defer graph DB until Phase 3.4. Qdrant alone covers ~80% of the assistant's daily value. The graph DB is justified once you actually need dependency analysis or critical-path queries.
Auto-export n8n workflows. nexa-core/scripts/backup_workflows.sh already exists. Schedule it inside the n8n container (cron) and let it git commit && git push — this is the cheapest disaster recovery.

For the wider homelab (out of scope but worth noting)

AI gateway naming. ai.nuclide.systems currently proxies LobeHub (a chat UI on :3210), while the LiteLLM API lives on :4000. For Nexa, point n8n directly at LiteLLM (http://192.168.1.40:4000 over the docker net — no public TLS hop needed) to save latency and isolate from UI restarts.
MCP servers consolidation. Dozzle shows crawl4ai-mcp, markitdown-mcp, papersearch-mcp running individually. They're all MCP servers — Nexa Phase-3 could pull from these via LiteLLM's MCP support to enrich the embedding pipeline (e.g. fetch + markitdown a Karakeep link before embedding).
Backup the n8n SQLite file — Backrest covers /home/node/.n8n if added; today the only "backup" is the workflow JSON which omits credentials and execution history.
Pocket-ID SSO in front of n8n would let you remove n8n basic-auth and unify session management across the whole stack. One-time setup, large UX win.
Vaultwarden as the secret store for Nexa secrets (SAIA_API_KEY, MEMOS_API_KEY, …) — read at bootstrap via the Bitwarden CLI from inside the docker host. Removes the need for a .env on disk.
AdGuard as DNS-based control plane. Since AdGuard is the resolver for the LAN, you can rewrite *.nuclide.systems to 192.168.1.4 (Zoraxy) internally and avoid a hairpin via the WAN — already the case if AdGuard rewrite rules are set, worth verifying.
Disk usage on LXC 104 is 47.7 % (Proxmox). Monitor; n8n execution logs and Dozzle history are the usual culprits. Setting EXECUTIONS_DATA_PRUNE=true and EXECUTIONS_DATA_MAX_AGE=168 (7 days) on n8n keeps it bounded.
Vector-store sprawl in the homelab — three competing indexes today. Nexa is about to be the fourth. Track for eventual consolidation:
- Obsidian Copilot (.copilot, .copilot-index) and Smart Connections / Smart Composer (.smart-env, ~13 MB) — embed the Obsidian vault into two separate vector stores inside the vault.
- Paperless-AI ChromaDB (services/paperless-ai/chromadb/) — embeds scanned documents from Paperless-ngx for Q&A.
- Nexa Qdrant (nexa_knowledge_text, soon _visual) — embeds the cross-source corpus.
Long-term, Nexa is the natural single source of truth (it sees memos + mail + obsidian + RDF graph). Once its RAG is satisfying, retire the Obsidian-plugin indexes and consider letting Nexa read from Paperless-AI's ChromaDB rather than re-embed PDFs (one-line ChromaDB query, much cheaper than redoing OCR-to-vector). Track but don't act yet.
Retire open-webui. Confirmed stale by the user — only LobeHub is in active use as the LiteLLM chat front-end (ai.nuclide.systems). Stop the container, tar services/open-webui/ into backup/open-webui/, then remove the stack. Frees ~500 MB RAM + a couple of GB of model cache.
services/siyuan/workspace/ is dead data. SiYuan retired, content migrated to Obsidian (Nextcloud Notizen/). Keep a final tar in backup/siyuan/, then rm -rf services/siyuan/. Frees disk + removes a "is this still authoritative?" question for future agents (and for Nexa's classifier if it ever sees the path).
Real-time Obsidian sync via notify_push. Phase 3.1 polls WebDAV every 15 min (Q4 resolution). Once that works, swap to Nextcloud's notify_push app for sub-second propagation. One-line workflow change in n8n.
assets/ is 186 MB of binaries in the Obsidian vault — worth a glance to confirm it's mostly images (Phase-3.2 visual queue) rather than something that should live in Nextcloud Files proper.
Paperless-AI as a Phase-2.x triage helper for _sortMe/Downloads/. UNAS shows ~230 PDFs/docs in _sortMe/Downloads/ plus another batch under _sortMe/Anne/. Paperless-AI (already running) can ingest, OCR, classify and route them; Nexa's role is to delegate — wire a workflow that posts a batch to Paperless-AI and reports the result back via Memos.
media/Recipes/ has ~300 individually-named recipe folders. Once Phase-3.1 text indexing works, this becomes a high-quality test corpus for #nexa:ask (e.g. "what was that gochujang noodle recipe with no anchovies?"). Out-of-scope for the deployment plan, but a satisfying first user-facing win.

Proxmox host (NUC 14 Pro) tuning

Observed from the node summary: 22 threads, 62 GiB RAM (32 GiB used, ~24 GiB of which is ZFS ARC), 1.64 TiB disk (0.35% used), load avg <2.0, IO delay 0.04%, kernel 6.17.13-4-pve, PVE 9.1.9, EFI. Suggestions in priority order:

Cap ZFS ARC. Default is 50% of RAM (~31 GiB); current actual ~24 GiB. For a node that runs services rather than a pure storage box, capping at 8–12 GiB frees ~12–16 GiB for guests without measurable IO impact (disk is 1.64 TiB and 0.35% used — there's nothing hot to cache):
```
echo 'options zfs zfs_arc_max=8589934592' > /etc/modprobe.d/zfs.conf   # 8 GiB
update-initramfs -u
```
Reboot or echo 8589934592 > /sys/module/zfs/parameters/zfs_arc_max to apply live.
Enable KSM (Kernel Same-page Merging). With ~40 docker containers + several LXCs, KSM typically frees 1–3 GiB by deduplicating identical memory pages. Currently KSM sharing: 0 B in the summary. PVE has ksmtuned available — systemctl enable --now ksmtuned.
Suppress the pve-no-subscription repository warning — either accept it (it's a homelab) and apply the pve-no-subscription-warning polyfill, or move to the enterprise repo. Pure cosmetic, but the orange banner in the UI is noise.
Swap is 31 GiB on a 62 GiB box with ZFS root — almost certainly oversized. Drop vm.swappiness to 10 (sysctl -w vm.swappiness=10 + persist) so swap is only used under genuine pressure, and consider shrinking the swap volume if disk-layout permits.
Verify scheduled ZFS scrub is enabled. PVE ships zfs-scrub-monthly@.timer — systemctl list-timers | grep zfs to confirm. Cheap insurance on a 1.6 TiB pool.
SMART monitoring on the NVMe. smartctl -a /dev/nvme0 should be regularly polled; PVE's notification target can ntfy on degradation. Combine with the existing nexa.system ntfy topic (optimization #4) so disk-health alerts land in the same Memos system feed as everything else.
NTP source via AdGuard. AdGuard already resolves DNS for the LAN; pointing the host's systemd-timesyncd at pool.ntp.org resolved through AdGuard avoids any external dependency for time. One-line change in /etc/systemd/timesyncd.conf.
fstrim.timer enabled for the SSD pool — verify with systemctl status fstrim.timer. Default-on in modern PVE, but quick to confirm.
Watchdog config is irrelevant for a single-node setup (HA is the use case), so leave the default. Mentioned only so future agents don't add it speculatively.

Docker LXC (104) — observations & wins

Confirmed allocation: 16 CPU, 31.25 GiB RAM (7.86 GiB used / 25%), 8 GiB swap (idle), 200 GiB boot disk at 47.7% used — disk pressure outranks RAM pressure.

Fix the Intel iGPU passthrough. The container's own notes flag the binding as "likely failing". Once /dev/dri/{card0,renderD128} is visible inside the LXC, both TEI and infinity can run embeddings on the Arc iGPU via OpenVINO / IPEX-LLM — typically 5–10× faster than CPU. Equally, Immich's CLIP can be GPU-accelerated. The required config in /etc/pve/lxc/104.conf:
```
lxc.cgroup2.devices.allow: c 226:0 rwm
lxc.cgroup2.devices.allow: c 226:128 rwm
lxc.cgroup2.devices.allow: c 29:0 rwm
lxc.mount.entry: /dev/dri/card0       dev/dri/card0       none bind,optional,create=file
lxc.mount.entry: /dev/dri/renderD128  dev/dri/renderD128  none bind,optional,create=file
lxc.idmap: u 0 100000 65536
lxc.idmap: g 0 100000 65536
lxc.idmap: g 44 44 1            # video group on host
lxc.idmap: g 104 104 1          # render group on host
```
Then inside the container: usermod -aG video,render <docker-user> and run TEI with --device cuda replaced by the OpenVINO build (text-embeddings-inference:cpu-1.5-openvino). Not needed for Phase-3.1's volume but a clean upgrade path.
Match the homelab-wide UNAS storage convention — host bind-mount of /mnt/pve/unas/services/<svc>/<vol>. Verified via the running Karakeep stack (Q19): every container in LXC 104 binds its persistent data into /mnt/pve/unas/services/<svc>/... directly, no driver_opts, no CIFS, no credentials in compose. Nexa follows the same pattern. Targets:
- /mnt/pve/unas/services/nexa/qdrant/ — Qdrant data dir.
- /mnt/pve/unas/services/nexa/tei-cache/ — embedding model cache (keeps multi-GB models off the boot disk).
- /mnt/pve/unas/services/nexa/graphdb/ — Phase-3.4 RDF store.
- /mnt/pve/unas/backup/nexa/snapshots/qdrant/<date>/ — daily Qdrant snapshots before rsync to S3 (Phase 3.3 — Q12 open). Mirrors the backup/<svc>/ pattern used by Home Assistant / Immich.
- n8n executions retention: EXECUTIONS_DATA_PRUNE=true, EXECUTIONS_DATA_MAX_AGE=168.
- Monthly docker image prune --all --filter "until=720h" to clear dangling layers.
Pre-deploy step: mkdir -p /mnt/pve/unas/services/nexa/{qdrant,tei-cache,graphdb} and mkdir -p /mnt/pve/unas/backup/nexa/snapshots/{qdrant,graphdb} on the docker LXC. Then the compose volumes block is just - /mnt/pve/unas/services/nexa/<vol>:/<container-path>. Concrete example in 09-deployment §Step 3.

Secrets go in the stack's .env next to the compose (Karakeep convention: env_file: .env). Vaultwarden is the long-term source-of-truth for those secrets (12/#11) but isn't required day-one. The earlier proposal of mounting SMB as a docker volume is withdrawn (12/#27) — it was an over-fitting to the Nextcloud-specific NFS issue.
The LXC has its own 8 GiB swap. Combined with the host's 31 GiB, that's a lot of swap for guests that should never page. Drop the LXC swap allocation to 1–2 GiB (pct set 104 -swap 2048) — frees disk on the LVM-thin pool and forces issues to surface earlier rather than silently swap.
Stacks are managed in Arcane; Dockge is stale. The LXC was originally provisioned with the Dockge helper-script template, but the user moved on to Arcane as the day-to-day docker manager. Deployment of the Nexa stack therefore goes through Arcane, not Dockge. Cleanup task: tar services/dockge/ (if it exists) into backup/dockge/, retire the Dockge container, drop the stale data dir. Dozzle stays — it's the log viewer, not a manager, so it isn't redundant with Arcane.
Untriaged local mail on the LXC. Console shows You have new mail. at login — the system mail spool on /var/mail/root has unread messages, almost always cron job failures. mailx or mutt to inspect, then either fix the failing job or send the spool to nexa.system ntfy via a tiny aliases entry (root: |/usr/local/bin/spool-to-ntfy.sh).

Housekeeping campaign (cross-cutting, do together)

These three are inter-related — picking them up as one campaign is cheaper than chasing each individually, because the audit step is the same. They also map cleanly onto Phase 6 (Nexa as homelab steward): once Nexa can poll Arcane, diff against the documented state and emit actionable findings, items #35–37 become semi-automatic — Nexa proposes the migrations rather than us hunting them down.

Consolidate Postgres instances. Today there are at least four independent Postgres containers running — visible from Arcane: immich_postgres, lobe-postgres, litellm_db, paperless-ngx-db-1 (Karakeep uses Meilisearch + maybe SQLite, separate). Each idles around 100–300 MB RAM and has its own backup story. Two paths:
- Single shared cluster (pgsql container with one role per app, one database per app). Most modern apps support DATABASE_URL → just point them at the shared instance. Saves ~600 MB–1 GB RAM and consolidates backups to one pg_dump cron.
- Or migrate the smallest ones to SQLite (LiteLLM has SQLite mode; Paperless does too) and keep Postgres only for Immich. Same outcome, less migration risk.
Pre-step: list every running container with docker ps --format '{{.Names}}\t{{.Image}}' | grep -i 'postgres\|mariadb\|mysql' to inventory exactly what's running.
UNAS-integration audit. /mnt/pve/unas/services/<svc>/ is the homelab convention (host bind-mount, no driver opts — see #31). Not every container follows it yet. Walk every Arcane stack and check the volumes: block: any - /var/lib/docker/... or anonymous-volume entry is non-compliant. Suspected non-compliant (need verification):
- qdrant_scientific — vector data possibly on local boot disk; critical to verify before Phase-3.1 piles on a nexa_knowledge_text collection.
- litellm_db, lobe-postgres, lobe-redis — DB containers, persistent state.
- memos — the user-facing notes app; loss = data loss.
- n8n — workflows + executions DB.
- audiobookshelf — listening progress + library metadata.
- the *-mcp containers — likely stateless (cache only), low priority.
Output: a one-page table service | persistent? | currently bound to | should be bound to. Then migrate the non-compliant ones one-by-one (stop → rsync data to services/<svc>/ on UNAS → re-create stack with the new bind → verify → keep the old volume for 7 days as a rollback). Lock the convention for any new stack going forward.
3-2-1 backup tiering: warm on-site (S3) + cold off-site. s3.nuclide.systems is already up but it's on-site — same building, same power, same (in-)susceptibility to fire / theft / ransomware. By itself it's a warm tier, not a disaster-recovery copy. The full design is two tiers:

Warm tier — s3.nuclide.systems (already exists, just needs a bucket):
- Restic / borg repos for backup/home-assistant/, backup/immich/, backup/nextcloud/, future backup/nexa/. Backrest already orchestrates Borg — just add the S3 destination.
- Qdrant + GraphDB snapshots flow UNAS → S3 daily.
- Resolves Q12 once the bucket name + access key are set.
Cold tier — off-site provider (decision pending — see Q20):
- Jottacloud "Unlimited" (~€9.50/mo, EU/Norway, soft-cap ~5 TB) — user's stated candidate. Best price/value at current 2 TB; reach soft cap in ~10 years. Use rclone jottacloud: or native jotta-cli.
- Hetzner Storage Box BX21 (€13/mo, 5 TB EU/DE) — predictable quota, native Borg/Restic/SFTP. Cheapest predictable EU alternative.
- Backblaze B2 (~~$12/mo for 2 TB, S3-compatible, US) — cheapest with the widest tooling support; egress is paid (~~$10/TB) which only bites during full restores.
- rsync.net (~$30/mo) — ZFS send/recv natively; overkill unless you want pool replication.
- Storj DCS ($4/TB/mo, decentralized, S3-compatible) — newer ecosystem.
Always encrypt before upload regardless of provider — restic or rclone crypt over the chosen remote. The provider sees only ciphertext blobs. Key material lives in Vaultwarden + a printed-paper offline copy.

Tiering & cadence:
```
[Source]   UNAS RAID-6 + native snapshots (#38)
             │ daily restic/borg
             ▼
[Warm]     s3.nuclide.systems (on-site, fast restore)
             │ weekly rclone copy + crypt
             ▼
[Cold]     Off-site (Jottacloud / B2 / Hetzner)  ← satisfies the "1" in 3-2-1
```
What goes off-site (priority order):
1. Immich photo originals — irreplaceable.
2. media/documents/, _sortMe/Downloads/, Nextcloud user data — irreplaceable.
3. Memos DB, n8n workflows, Nexa Qdrant + GraphDB snapshots — replicable from sources but expensive to redo.
4. Vaultwarden DB — cryptographically sensitive but small; explicitly include.
What does not need off-site: movies, music, audiobooks, ROMs, derived caches (Immich thumbnails, encoded-video). Those re-download or regenerate.

Pre-deploy step: choose one provider, create a test bucket/folder, dry-run a restic init + restic backup of backup/nexa/ first as a smoke test before pointing the heavyweight repos at it.
Configure UNAS Pool snapshots. The UniFi Drive dashboard shows the storage-pool snapshot schedule as "Click to Setup" — i.e. not configured. RAID 6 protects against drive failure, not against an rm -rf from a misbehaving container or a Nextcloud user mass-delete. Cheap fix: set a daily snapshot with 7-day retention + a weekly with 4-week retention on Pool 1. Native to UniFi Drive, no agent needed. Highest-leverage data-protection change in the homelab right now — costs nothing, recovers everything.
Use UNAS native snapshots for fast Nexa rollback once #38 is set. Nexa workflows that mutate large state (re-embedding the whole vault, graph rebuild, list-config rewrite) can pre-snapshot → operate → verify → release against the same UniFi Drive snapshots. Cheaper and faster than restoring from S3.

19 KiB Raw Blame History Unescape Escape