Files
bluejay-infra/apps/intranet/intranet.yaml
Andrew Stoltz 37ce0aed85 intranet: v202604240135longchunk — long-chunk handling fix
Image bump v202604240108gpu -> v202604240135longchunk, rebuilt from
FlowerCore.Intranet.Web@feat/shared-indexing-search HEAD which transitively
picks up FlowerCore.Common@feat/shared-indexing@105af75:

- MarkdownChunker hard-caps oversized heading-bounded sections at
  ChunkSizeTokens × 4 chars and splits with overlap (same pattern as
  JsonArticleChunker). Stops the indexer from producing chunks above
  nomic-embed-text's 8192-token input limit at the source.

- IndexBuilder gains IndexingOptions.MaxEmbeddingTokens (default 8000)
  safety filter — chunks above the cap are warn-logged and dropped
  before any batch is sent. New IndexBuildResult.ChunksDropped tracks
  how many got skipped.

Goal: notes-md should index 2541/2541 chunks (vs. 2080/2541 last pass)
with zero "Failed to embed batch" 400s.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-24 01:28:00 -05:00

124 lines
2.7 KiB
YAML
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
apiVersion: v1
kind: Namespace
metadata:
name: intranet
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: intranet-vector-store
namespace: intranet
spec:
accessModes:
- ReadWriteOnce
storageClassName: longhorn
resources:
requests:
storage: 1Gi
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: intranet-web
namespace: intranet
labels:
app: intranet-web
spec:
replicas: 1
strategy:
type: Recreate
selector:
matchLabels:
app: intranet-web
template:
metadata:
labels:
app: intranet-web
spec:
containers:
- name: intranet-web
image: localhost/fc-intranet-web:v202604240135longchunk
imagePullPolicy: Never
ports:
- containerPort: 5300
name: http
env:
- name: ASPNETCORE_ENVIRONMENT
value: Production
- name: ASPNETCORE_URLS
value: "http://+:5300"
# Bulk corpus indexing on edge1 Pi 5 takes ~6s/chunk × 5665 chunks
# ≈ 9 hours. BLUEJAY-WS GPU (R9700, 32GB VRAM) does the same work
# in minutes. Memory: feedback_pi5_nomic_embed_slow.
- name: IntranetSearch__OllamaBaseUrl
value: "http://10.0.56.20:11434"
resources:
requests:
memory: "256Mi"
cpu: "100m"
limits:
memory: "1Gi"
cpu: "1000m"
livenessProbe:
httpGet:
path: /health
port: 5300
initialDelaySeconds: 30
periodSeconds: 30
readinessProbe:
httpGet:
path: /health
port: 5300
initialDelaySeconds: 10
periodSeconds: 10
volumeMounts:
- name: vector-store
mountPath: /data
volumes:
- name: vector-store
persistentVolumeClaim:
claimName: intranet-vector-store
---
apiVersion: v1
kind: Service
metadata:
name: intranet-web
namespace: intranet
spec:
selector:
app: intranet-web
ports:
- port: 5300
targetPort: 5300
name: http
---
apiVersion: cert-manager.io/v1
kind: Certificate
metadata:
name: intranet-tls
namespace: intranet
spec:
secretName: intranet-tls
issuerRef:
name: step-ca-acme
kind: ClusterIssuer
dnsNames:
- intranet.iamworkin.lan
---
apiVersion: traefik.io/v1alpha1
kind: IngressRoute
metadata:
name: intranet
namespace: intranet
spec:
entryPoints:
- websecure
routes:
- match: Host(`intranet.iamworkin.lan`)
kind: Rule
services:
- name: intranet-web
port: 5300
tls:
secretName: intranet-tls