Files
homelab/cluster/applications/hindsight/externalsecret.yaml
Hermes Agent service account 2cc9370f3d deploy-vllm: add embedding-mode support, cut over Hindsight to vLLM (t_e6facb19)
- vllm.service.j2: branch on role==embedding for --runner pooling
  --convert embed, --no-enable-prefix-caching, per-model
  trust_remote_code toggle (needed for nomic-embed-text-v1.5's custom
  NomicBertModel code), and enforce_eager toggle (needed to avoid CUDA
  graph capture OOM when co-resident with another vLLM process on this
  24GB card).
- tasks/verify.yml: split completions vs embedding smoke tests --
  embedding-mode instances don't serve /v1/completions. Assert a
  non-empty embedding vector, not just HTTP 200.
- host_vars/astro-orbiter: enable nomic-embed-text-v1.5 (port 8020),
  lower primary model's gpu_memory_utilization 0.95->0.90 + add
  enforce_eager after finding 0.95 crash-looped 6-7x before stabilizing
  with co-resident nomic-embed (real fix, confirmed via NRestarts=0
  after clean stop/start, not luck).
- Hindsight (values.yaml + externalsecret.yaml): cut LLM + embeddings
  over to vLLM (:8000, :8020), wire the previously-unset
  HINDSIGHT_API_EMBEDDINGS_* env vars for the first time, and swap the
  API key secret source from the Nous fallback item to vllm/api-key
  (vLLM enforces real auth, llama-swap did not).
- README: document the embedding-mode branch, VRAM findings, and a
  genuine architecture gap -- vLLM's one-model-per-process design
  cannot replace llama-swap's 5-model LRU roster on this 24GB card, so
  21 Hermes profiles' aux-model consumers (Qwen3-8B-no_think,
  Phi-3.5-mini, Meta-Llama-3.1-8B, Qwen2.5-Coder-14B) and OpenViking's
  VLM stay on llama-swap. Full teardown (t_6dff1ecc) needs a human
  decision on the aux-model strategy before it can proceed.
2026-08-31 18:15:22 -05:00

93 lines
4.7 KiB
YAML

# ============================================================================
# ExternalSecret - Hindsight Credentials
# Peter Parker, Phase C, t_6d47a360
#
# Syncs Hindsight credentials from the 1Password `mk-labs` vault (item
# "hindsight", UUID q6pwoohexogdyvitt7sm2xcusu).
# Pattern: openviking proven pattern (cluster/platform/openviking/externalsecret.yaml).
# Store: onepassword-connect ClusterSecretStore.
# Namespace: hindsight.
#
# Wave: -1 — must sync BEFORE the chart's Deployments (which default to wave 0)
# so the materialized Secret exists before the api / control-plane pods attempt
# envFrom / secretKeyRef injection.
#
# The materialized Secret (hindsight-credentials) must carry keys in TWO shapes
# (Phase B record: inbox/ryan/2026-08-24-hindsight-phase-b-secrets.md):
# 1. `postgres-password` (lowercase, hyphenated) — read by the chart's EXPLICIT
# secretKeyRef for POSTGRES_PASSWORD, NOT via envFrom. A naming error here
# silently drops the DB password (pod starts, then fails to connect).
# 2. `HINDSIGHT_API_LLM_API_KEY` / `HINDSIGHT_API_MCP_AUTH_TOKEN` — env-var-named,
# injected via envFrom (always runs when existingSecret is set).
#
# 1Password item "hindsight" fields (all CONCEALED):
# - postgres-password (32 chars, letters+digits only, URL-safe)
# - HINDSIGHT_API_MCP_AUTH_TOKEN (48 chars urlsafe bearer token)
# - HINDSIGHT_API_LLM_API_KEY (literal "local-placeholder" — astro-orbiter
# does not validate; must rotate if OpenRouter
# hosted fallback is ever enabled)
# ============================================================================
apiVersion: external-secrets.io/v1beta1
kind: ExternalSecret
metadata:
name: hindsight-credentials
namespace: hindsight
annotations:
# Wave -1: ensure the ExternalSecret syncs BEFORE the chart's Deployments
# (wave 0). Guarantees the Secret exists before the api / control-plane
# pods attempt envFrom / secretKeyRef injection.
argocd.argoproj.io/sync-wave: "-1"
description: "Phase C secrets for Hindsight deployment (1Password mk-labs item: hindsight)"
spec:
refreshInterval: "1h"
secretStoreRef:
kind: ClusterSecretStore
name: onepassword-connect
target:
name: hindsight-credentials
creationPolicy: Owner
template:
engineVersion: v2
data:
# LOWERCASE, HYPHENATED — read by the chart's explicit secretKeyRef for
# POSTGRES_PASSWORD. Must be exactly "postgres-password".
postgres-password: "{{ .postgresPassword }}"
# env-var-named keys injected via envFrom
HINDSIGHT_API_MCP_AUTH_TOKEN: "{{ .HINDSIGHT_API_MCP_AUTH_TOKEN }}"
HINDSIGHT_API_LLM_API_KEY: "{{ .HINDSIGHT_API_LLM_API_KEY }}"
# Cutover to vLLM (t_e6facb19, 2026-08-31): vLLM enforces its API key
# on every request (unlike llama-swap, which accepted any/no key) —
# confirmed empirically, a bad/missing key gets a real 401
# {"error":"Unauthorized"}. Both the completions endpoint (:8000) and
# the embeddings endpoint (:8020) are separate vLLM processes but
# share the SAME key (api-key.env is written once, read by both
# systemd units per roles/deploy-vllm/tasks/api-key.yml). Reuse
# HINDSIGHT_API_LLM_API_KEY's value for the embeddings key too.
HINDSIGHT_API_EMBEDDINGS_OPENAI_API_KEY: "{{ .HINDSIGHT_API_LLM_API_KEY }}"
data:
# 1Password item "hindsight", field "postgres-password".
# letters+digits only / URL-safe: required because the chart interpolates the
# value RAW into HINDSIGHT_API_DATABASE_URL (no URL-encoding).
- secretKey: postgresPassword
remoteRef:
key: hindsight
property: postgres-password
# 1Password item "hindsight", field "HINDSIGHT_API_MCP_AUTH_TOKEN"
# (MCP bearer token, D6 — enables bearer-auth on the /mcp endpoint)
- secretKey: HINDSIGHT_API_MCP_AUTH_TOKEN
remoteRef:
key: hindsight
property: HINDSIGHT_API_MCP_AUTH_TOKEN
# Cutover to vLLM (t_e6facb19, 2026-08-31): astro-orbiter's llama-swap
# (which needed no real auth) is being retired for Hindsight's
# completions+embeddings roles. vLLM enforces a real API key — read the
# SAME key roles/deploy-vllm's api-key.yml phase writes to
# /etc/vllm/api-key.env on astro-orbiter, sourced from 1Password
# op://mk-labs/vllm/api-key (item "vllm", field "api-key", vault mk-labs).
# Replaces the prior "nous" item's api-key (Nous free-tier fallback,
# no longer the active LLM backend once this cutover lands).
- secretKey: HINDSIGHT_API_LLM_API_KEY
remoteRef:
key: vllm
property: api-key