Previous comments described llm_ttft_seconds / llm_itl_seconds /
llm_tokens_total as "streaming-only". That was wrong, and optimistic.
They are dead code. internal/proxy/llm_metrics.go defines
LLMMetricsCapture, which is the only thing that calls RecordTTFT /
RecordITL / RecordTokens — but NewLLMMetricsCapture() has zero call sites
anywhere in the repo, verified by grep. The exporter skips keys with no
samples, so those three series render as HELP-only stubs and can never be
graphed, streaming or not.
No behaviour change here; only the comments and the dashboard generator's
header are corrected so the next person does not build on a false
assumption. The gateway latency panels are left in place so the gap stays
visible — they populate unchanged once the capture wrapper is wired into
the streaming proxy path.
Series that DO work from the gateway, confirmed live with 3 series (one
per replica): queue_full_requests_total, warmup_requests_delayed_total,
vllm_prefix_cache_hit_tokens_total.
Both new ServiceMonitors sat permanently OutOfSync. Cause: `action: replace`
is the API server default for relabeling rules, so the live object carried it
while the manifests omitted it.
Rather than write out the default to paper over drift, removed the rules that
were never needed:
- vllm.yaml: the namespace/pod/service relabelings were redundant —
Kubernetes service discovery already supplies all three. Only `app` has to
be derived (TEI panels aggregate by it and SD does not promote arbitrary
pod labels). Kept that one, with an explicit action.
- api-gateway.yaml: the metricRelabelings block was a no-op — it mapped
__address__ onto instance, which is already the default. Removed. The
series inventory it was documenting moved up into the file header, which
is where it belonged anyway.
New Grafana dashboard (llm-observability) with 27 panels across 5 rows:
TTFT, inter-token latency, batching/scheduling, KV cache, throughput.
Both metric sources are shown side by side because they answer different
questions and neither is complete alone:
vLLM 0.11.0 (reasoning) — engine truth: kv_cache_usage_perc,
num_requests_running/waiting, num_preemptions_total, prefill vs
decode time, prefix cache, per-request histogram quantiles.
TEI 1.8.2 (embeddings, reranker) — te_batch_next_size, te_queue_size,
te_embed_duration, te_embed_queue_duration.
api-gateway — the ONLY per-request latency source covering Ollama
models (ornith:35b, qwen2.5:3b-instruct), which expose no metrics.
It is derived from the SSE byte stream, so it is streaming-only and
counts SSE chunks rather than true tokens. vLLM's own histograms are
authoritative for reasoning.
All 44 panel queries were executed against live Prometheus before commit —
all return success. vLLM/TEI panels have live series; gateway panels show
no data until the netpol fix in homelab-frontend PR #62 merges.
Also fixes a dead alert. alerts/cag-kv-cache-rules.yaml derived
vllm:prefix_cache_hit_rate from vllm:cache_query_hit_total and
vllm:cache_query_miss_total. Those metric names do not exist in vLLM
0.11.0 — the real series are vllm:prefix_cache_hits_total and
vllm:prefix_cache_queries_total. The recording rule produced no data, so
VLLMPrefixCacheHitRateLow could never fire.
Not included: real GPU hardware metrics (SM utilisation, framebuffer,
power, temperature). Only nvidia-device-plugin is deployed, which exposes
none of them; DCGM exporter is required. vllm:kv_cache_usage_perc and
preemption rate are the only GPU-pressure proxies available today.
Both metric sources existed but neither was scraped:
- servicemonitors/vllm.yaml was orphaned (absent from kustomization
resources:) and would not have matched anyway — it used 'port: http'
but KServe-generated Services in llm-serving have UNNAMED ports.
Switched to targetPort: 8080 (prometheus-operator v0.94.1 supports it),
which avoids editing the InferenceServices and therefore avoids a
predictor restart / weights reload.
- api-gw exposes llm_ttft_seconds, llm_itl_seconds, llm_tokens_total,
queue_full_requests_total, warmup_requests_delayed_total and
vllm_prefix_cache_hit_tokens_total, but the Deployment's
prometheus.io/scrape annotation is ignored by kube-prometheus-stack
(no additionalScrapeConfigs). Added a ServiceMonitor for it. This is
the only per-request latency source covering the two Ollama models,
which expose no metrics endpoint at all.
Ollama services are excluded from the vllm ServiceMonitor selector to
avoid permanently-DOWN targets.
Note: alerts/cag-kv-cache-rules.yaml references vllm:prefix_cache_hit_rate
derived from vllm:cache_query_hit_total/miss, which were never scraped.
Those alerts were dead; verify the metric names after the first scrape.
Actual TSDB data is ~2Gi. 100Gi over-provisioned Longhorn storage
contributed to the cluster-wide scheduling pressure; 25Gi is ample
with room for retention growth.
GET /user/sign_up returned 200 — anyone reaching forgejo.riotpiao.com
could create an account. Add [service] DISABLE_REGISTRATION: true.
Admin-created accounts (UI/API) are unaffected.
32 degraded + 3 faulted volumes from replica rebuild failures.
cp-1/cp-2 over-committed; worker-1 had 602Gi free but scheduling
disabled. Enable it with 400Gi reserved for LLM weights.
- Add worker-1 to longhorn-nodes.yaml (allowScheduling: true)
- Add worker-1 disk config to add-disks-job (storageReserved: 400Gi,
evictionRequested: false)
GITEA__WEBHOOK__ALLOWED_HOST_LIST=external,api.riotpiao.com
api.riotpiao.com resolves to cluster ingress ClusterIP (private range)
which is blocked by Forgejo's default SSRF protection. Safe to allowlist
as it is our own api-gw.
Points to riotpiao-poimen/poimen-memory.git k8s/app/ (kustomize).
ArgoCD auto-syncs Redis deployment, NetworkPolicies, ConfigMap, Service.
Image updater: digest strategy on :latest tag for rolling deployments.
Wave 7 — same wave as poimen workflow app.
V100-PCIE-32GB has ~12GB VRAM headroom after model weights.
ornith:35b Q4_K_M KV cache at 64K ≈ 11.8GB — fits within budget.
Model natively supports 256K but hardware caps safe limit at 64K.
nginx-ingress v1.15.1 blocks configuration-snippet by default as a
security risk, silently dropping the entire Ingress resource. WebSocket
upgrade headers are handled natively with proxy-http-version 1.1.
- Increase PVC from 20Gi to 100Gi (talos-cp-1 disk full)
- Move from az-a (talos-cp-1) to worker-1 (648GB available)
- Increase retention from 15d to 30d (95GB limit)
- Fixes Prometheus pod restarts due to disk pressure
- Add MINIO_BROWSER_REDIRECT_URL pointing to public URL
- Add MINIO_BROWSER_LOGIN_ANIMATION=off
- Enables Console to show OIDC/Authentik SSO button on login page
- Change MINIO_IDENTITY_OPENID_CLAIM_NAME from 'policy' to 'groups'
- Add MINIO_IDENTITY_OPENID_CLAIM_PREFIX (empty) to use raw group names
- Update scopes to 'openid,profile,email,groups'
- Enables Authentik group-based access (homelab-admins, minio-admins, etc)
- Users in homelab-admins group will get admin access to MinIO
The correct env var is PAPERLESS_SOCIAL_AUTO_SIGNUP not PAPERLESS_SOCIALACCOUNT_AUTO_SIGNUP
This enables automatic linking of Authentik social accounts to existing Paperless users
- Add SOCIALACCOUNT_AUTO_SIGNUP for automatic social account creation
- Enable ACCOUNT_ALLOW_SIGNUPS for OIDC users
- Add SOCIALACCOUNT_EMAIL_AUTHENTICATION_AUTO_CONNECT to link existing users
- Fixes 'An account already exists with this email' error
- Allows Authentik login to auto-link to existing Paperless users
Re-enable skip_org_role_sync=false to sync Admin role from grafana-admins group.
User creation worked with skip_org_role_sync=true, now restore role sync.
Set skip_org_role_sync=true to allow user creation on first OAuth signin.
When false, Grafana tries to sync org roles before user exists, causing creation to fail.
Reverts to standard OpenID 'preferred_username' claim which is guaranteed to be present in all userinfo responses. The 'email' claim requires explicit scope mapping in Authentik that may not be consistently returned.
Authentik user 'rock' has preferred_username='rock' which matches the existing Grafana user login.
- Commit paperless-ai-config.yaml with production environment variables
- Configure Paperless API endpoint and token
- Enable AI processing with qwen2.5:3b model
- Set up auto-tagging, correspondent/document type extraction
- Configure 60s scan interval for document processing
- Add ConfigMap to kustomization.yaml for GitOps deployment
- Add ConfigMap with production paperless-ai config
- Mount .env config file for app startup
- Enable auto-tagging, correspondent extraction, document type detection
- Set LLM API endpoint and token file location
- Configure 60s scan interval for document processing
- Change login_attribute_path from preferred_username to email for stable user matching
- Enable allow_sign_up to permit OAuth user sync with existing local users
- Root cause: Authentik's 'rock' user matches existing Grafana 'rock' by email, not by preferred_username claim
- Remove hardcoded client secret from init container
- Reference secret from paperless-ai-config (SOPS-encrypted)
- Init container reads LLM_AUTH_CLIENT_SECRET env var
- Secret key must be added to k8s/argocd/secrets/paperless-ai-secrets.enc.yaml
by someone with SOPS/age key access
This removes the plaintext secret from commit history and future
deployments will source it securely from the encrypted Secret.
- Add paperless namespace to llm-serving NetworkPolicy ingress
- Point paperless-ai CUSTOM_BASE_URL to reasoning-predictor directly
- Gateway JWT auth not yet available for service accounts
- paperless-ai-agent has llm:inference role in Authentik
- stale-job-cleanup: run every 2m, delete completed/failed jobs+pods after 2m
- paperless-ai: add PAPERLESS_USERNAME=admin (required to find own user ID)
ROOT CAUSE:
- tektoncd/pipeline config/ dir uses ko:// image refs (build-time placeholders)
- ArgoCD synced the raw dev manifests → InvalidImageName on all pods
- tektoncd/operator requires its own CRDs and controller (too heavy)
- Tekton has no official Helm chart
FIX:
- Vendor the pre-built release.yaml (v0.68.0) into k8s/infra/tekton/
- Point ArgoCD Application at our own repo (forgejo)
- Release contains real container images (ghcr.io/tektoncd/pipeline/*)
- Remove external tektoncd repo from AppProject (not needed)
TO UPGRADE TEKTON:
1. Download new release from github.com/tektoncd/pipeline/releases
2. Replace k8s/infra/tekton/release.yaml
3. Commit — ArgoCD syncs automatically
Required by paperless-ai to find its own user ID for scanning.
Without this, scanning aborts with "Failed to get own user ID".Reviewed-on: #45
Co-authored-by: rock <[email protected]>
Root cause: k8s Service named 'comfyui' auto-injects COMFYUI_PORT=tcp://...
into pod env, clobbering ai-dock's own COMFYUI_PORT variable which expects
a port number. This broke caddy's proxy config, leaving port 8188 dead.
Fix: enableServiceLinks: false, revert to port 8188 (ai-dock default).
Also bump startup probe failureThreshold 60->120 (20min budget for model loading).
- paperless-ai deployment using clusterzx/paperless-ai
- LLM via local api-gateway (reasoning model, no auth - phase 3 pending)
- Paperless API token SOPS-encrypted
- Auto-tags new documents, 5min scan interval
- Adds 'ai-processed' tag to classified documents
Liveness probe was firing after 60s with failureThreshold:6, killing
the server container before it finished applying 200+ DB migrations.
The startup probe (20min timeout) never got a chance to complete.
Root cause: Authentik health checks fail during long DB bootstrap.
Both liveness and startup probes run in parallel. Liveness killed the
pod at 60s; migrations need 2-3min minimum.
Solution: Add initialDelaySeconds:300 to liveness/readiness probes so
they don't fire until 5min have passed (migrations definitely complete).
Worker gets same treatment since it depends on server's DB bootstrap.
- Added queue OAuth2 provider (pk=13, client_id=queue-sqs)
- Added client_credentials grant type to all OAuth2 providers
- Updated temporal-worker-agent service account:
- Added queue:send role
- Added sqs_queues=* claim
- Provisioning script now creates 6 OAuth2 providers (all complete)
- All groups, scopes, and service accounts ready for Phase 2
Updated 2026-09-12 in provision-rbac.py
uid/gid are FAT/NTFS/FUSE mount options, not ext4. Caused mount exit 32
on every newly provisioned CNPG PVC, blocking authentik-db initdb.
CNPG handles postgres ownership via its own init container.
Add ServiceMonitor for llm-serving namespace. Wire prometheus.io annotations to all vLLM pods (reasoning, ornith, embeddings, reranker). Scrape /metrics@8080 every 30s with proper relabeling.
ArgoCD couldn't decrypt secrets.yaml because it was listed as a plain
kustomize resource. Move the 3 secrets (gotify-admin, gotify-tokens,
gotify-smtp) to k8s/argocd/secrets/ as .enc.yaml files processed by
the ksops generator, matching the repo convention.
Fixes ComparisonError: 'Object Kind is missing' (SOPS ciphertext
parsed as raw YAML).
Repos transferred: homelab-frontend, kmsvc-manage, poimen, poimen-memory,
poimen-workflows, poimen-frontend. Old URLs return 301 which ArgoCD
doesn't follow.
NOT changed: container image registry paths (rock/ is correct for registry),
riotpiao.com (still under rock org).
Also adds poimen-frontend to AppProject sourceRepos allowlist.
- Change container port from 8188 to 8888 (Caddy proxy binding)
- Remove PVC, use emptyDir for ephemeral models/output
- Replace startup + readiness probes with single liveness probe
- Remove explicit COMFYUI_FLAGS (container defaults work)
- Pod now reaches Ready state immediately after image pull
Fixes GPU contention by using ephemeral storage. ComfyUI now runs and is accessible at https://comfy.riotpiao.comReviewed-on: #18
Co-authored-by: poimen <[email protected]>
Reviewed-on: riotpiao-poimen/homelab-frontend#21
Co-authored-by: rock <[email protected]>
## Summary
Replaces hand-applied `llm-serving-default-deny` NetworkPolicy with a git-managed, namespace-scoped policy that only allows traffic from the api-gateway.
**Issue:** #13
Co-authored-by: rock <[email protected]>
- buildOptions field not available in ArgoCD v3.4.5
- SOPS decryption already handled by repo-server ksops plugin
- Revert to simple kustomize config
- Portfolio Application can now sync properly
- Add kustomize config with --enable-alpha-plugins to support ksops
- Allows ArgoCD to properly decrypt SOPS-encrypted files
- Fixes Image Updater compatibility with sops field in kustomization.yaml
Job containers spawned by the runner run inside dind. With TCP+TLS
(tcp://localhost:2376), localhost inside those containers doesn't
reach the dind sidecar. Unix socket at /run/docker.sock works because
both runner and dind share the /run emptyDir.
Also disables DOCKER_TLS_CERTDIR so dind creates the socket instead
of only listening on TLS TCP.
Previous image (golang:1.26-bookworm) lacks Node.js, causing GitHub Actions
to fail with: 'exec: "node": executable file not found'
Solution:
- Change golang runner image from golang:1.26-bookworm to docker:27-cli
- docker:27-cli includes: Node.js, Go toolchain, git, docker CLI, full dev tools
- Verified tag exists: docker manifest inspect docker:27-cli ✓
This allows actions/checkout@v4 and other GitHub Actions to run properly
on the golang runner pod.
Note: node:22-bookworm runner already has Node.js, no change needed.
Previous image (rust:1.83-bookworm) lacks Node.js, causing GitHub Actions
to fail with: 'exec: "node": executable file not found'
Solution:
- Change rust runner image from rust:1.83-bookworm to docker:27-cli
- docker:27-cli includes: Node.js, git, docker CLI, full dev tools
- Verified tag exists: docker manifest inspect docker:27-cli ✓
This allows actions/checkout@v4 and other GitHub Actions to run properly
on the rust runner pod.
- ClusterRoleBinding doesn't use spec: wrapper (unlike Deployment/StatefulSet)
- roleRef and subjects go at top level with metadata
- Fixes: 'strict decoding error: unknown field "spec"'
- RBAC kustomization contains cluster-scoped (ClusterRoleBinding) and
namespace-scoped (Role/RoleBinding) resources
- Each resource has explicit metadata.namespace, so Application shouldn't
force a default namespace
- Fixes: ClusterRoleBinding gets namespace=default, causing sync failure
with 'unsupported role reference kind: ""'
rock
