Add activity-core llm-connect reconcile gate

2026-06-18 02:10:18 +02:00
parent ff8cbe596c
commit 5f8d2831c7
3 changed files with 393 additions and 1 deletions
--- a/5
+++ b/5
@@ -27,6 +27,9 @@ test-ha-failover: ## Run HA failover test (D3) — kills primary PG pod, asserts
 verify-activity-core: ## Reconcile activity-core runtime and verify disabled ops inventory probe evidence
 	tools/cmd/railiance-verify-activity-core

+reconcile-activity-core-llm-connect: ## Reconcile activity-core llm-connect URL and run non-secret gate checks
+	tools/cmd/railiance-reconcile-activity-core-llm-connect
+
 ##@ Help

 help: ## Show this help
@@ -34,4 +37,4 @@ help: ## Show this help
 	  /^[a-zA-Z_-]+:.*?##/ { printf "  \033[36m%-20s\033[0m %s\n", $$1, $$2 } \
 	  /^##@/ { printf "\n\033[1m%s\033[0m\n", substr($$0, 5) }' $(MAKEFILE_LIST)

-.PHONY: backup restore preflight k3s-install smoke test-ha-failover verify-activity-core help
+.PHONY: backup restore preflight k3s-install smoke test-ha-failover verify-activity-core reconcile-activity-core-llm-connect help
--- a/tools/cmd/railiance-reconcile-activity-core-llm-connect
+++ b/tools/cmd/railiance-reconcile-activity-core-llm-connect
@@ -0,0 +1,293 @@
+#!/usr/bin/env bash
+# Cluster-owned activity-core <-> llm-connect reconcile and non-secret evidence.
+set -euo pipefail
+
+ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
+
+NAMESPACE="${ACTIVITY_CORE_NAMESPACE:-activity-core}"
+CLUSTER_HOST="${ACTIVITY_CORE_CLUSTER_HOST:-railiance01}"
+STATE_HUB_URL="${STATE_HUB_URL:-http://127.0.0.1:8000}"
+
+EXPECTED_URL="${LLM_CONNECT_URL:-http://llm-connect.activity-core.svc.cluster.local:8080}"
+EXPECTED_TIMEOUT="${LLM_CONNECT_TIMEOUT_SECONDS:-300}"
+SECRET_NAME="${LLM_CONNECT_PROVIDER_SECRET_NAME:-llm-connect-provider-secrets}"
+DEPLOYMENT_NAME="${LLM_CONNECT_DEPLOYMENT_NAME:-llm-connect}"
+LLM_CONNECT_REPO="${LLM_CONNECT_REPO:-/home/worsch/llm-connect}"
+LLM_CONNECT_REMOTE_REPO="${LLM_CONNECT_REMOTE_REPO:-}"
+APPLY_LLM_CONNECT_OVERLAY="${APPLY_LLM_CONNECT_OVERLAY:-1}"
+REQUIRE_SMOKE="${REQUIRE_LLM_CONNECT_SMOKE:-0}"
+
+EVIDENCE_WORKSTREAM_ID="${STATE_HUB_EVIDENCE_WORKSTREAM_ID:-}"
+EVIDENCE_TASK_ID="${STATE_HUB_EVIDENCE_TASK_ID:-}"
+
+PATCH_JSON="$(
+  EXPECTED_URL="$EXPECTED_URL" EXPECTED_TIMEOUT="$EXPECTED_TIMEOUT" python3 - <<'PY'
+import json
+import os
+
+print(json.dumps({
+    "data": {
+        "LLM_CONNECT_URL": os.environ["EXPECTED_URL"],
+        "LLM_CONNECT_TIMEOUT_SECONDS": os.environ["EXPECTED_TIMEOUT"],
+    }
+}))
+PY
+)"
+
+LIVE_URL=""
+LIVE_TIMEOUT=""
+SECRET_STATUS="unknown"
+SECRET_KEY_COUNT="0"
+DEPLOYMENT_STATUS="unknown"
+SMOKE_STATUS="skipped"
+SMOKE_SUMMARY=""
+EVIDENCE_STATUS="passed"
+FAILING_GATE=""
+
+export NAMESPACE CLUSTER_HOST STATE_HUB_URL EXPECTED_URL EXPECTED_TIMEOUT
+export SECRET_NAME DEPLOYMENT_NAME LLM_CONNECT_REPO LLM_CONNECT_REMOTE_REPO
+export APPLY_LLM_CONNECT_OVERLAY REQUIRE_SMOKE EVIDENCE_WORKSTREAM_ID EVIDENCE_TASK_ID
+export LIVE_URL LIVE_TIMEOUT SECRET_STATUS SECRET_KEY_COUNT DEPLOYMENT_STATUS
+export SMOKE_STATUS SMOKE_SUMMARY EVIDENCE_STATUS FAILING_GATE
+
+log() {
+  printf '[activity-core-llm-connect] %s\n' "$*"
+}
+
+quote() {
+  printf '%q' "$1"
+}
+
+cluster_bash() {
+  local script="$1"
+  if [[ -n "$CLUSTER_HOST" ]]; then
+    ssh "$CLUSTER_HOST" "bash -s" <<<"$script"
+  else
+    bash -s <<<"$script"
+  fi
+}
+
+post_evidence() {
+  python3 - <<'PY'
+import json
+import os
+import sys
+import urllib.request
+
+status = os.environ["EVIDENCE_STATUS"]
+detail = {
+    "producer": "railiance-cluster",
+    "verification": "activity-core llm-connect live reconcile",
+    "status": status,
+    "failing_gate": os.environ.get("FAILING_GATE") or None,
+    "cluster_host": os.environ.get("CLUSTER_HOST") or "local-kubectl",
+    "namespace": os.environ["NAMESPACE"],
+    "expected_url": os.environ["EXPECTED_URL"],
+    "expected_timeout_seconds": os.environ["EXPECTED_TIMEOUT"],
+    "live_url": os.environ.get("LIVE_URL") or None,
+    "live_timeout_seconds": os.environ.get("LIVE_TIMEOUT") or None,
+    "provider_secret": {
+        "name": os.environ["SECRET_NAME"],
+        "status": os.environ.get("SECRET_STATUS"),
+        "key_count": int(os.environ.get("SECRET_KEY_COUNT") or "0"),
+    },
+    "deployment": {
+        "name": os.environ["DEPLOYMENT_NAME"],
+        "status": os.environ.get("DEPLOYMENT_STATUS"),
+    },
+    "smoke": {
+        "status": os.environ.get("SMOKE_STATUS"),
+        "summary": os.environ.get("SMOKE_SUMMARY") or None,
+    },
+}
+
+if status == "passed":
+    summary = (
+        "Railiance activity-core llm-connect reconcile passed: runtime config, "
+        "provider Secret, deployment, and smoke gate are all healthy."
+    )
+elif status == "blocked":
+    summary = (
+        "Railiance activity-core llm-connect reconcile is blocked: "
+        f"{os.environ.get('FAILING_GATE') or 'operator gate'}."
+    )
+else:
+    summary = (
+        "Railiance activity-core llm-connect reconcile failed"
+        + (f" at {os.environ.get('FAILING_GATE')}" if os.environ.get("FAILING_GATE") else "")
+        + "."
+    )
+
+payload = {
+    "summary": summary,
+    "event_type": "note",
+    "author": "railiance-cluster",
+    "detail": detail,
+}
+if os.environ.get("EVIDENCE_WORKSTREAM_ID"):
+    payload["workstream_id"] = os.environ["EVIDENCE_WORKSTREAM_ID"]
+if os.environ.get("EVIDENCE_TASK_ID"):
+    payload["task_id"] = os.environ["EVIDENCE_TASK_ID"]
+
+body = json.dumps(payload).encode("utf-8")
+req = urllib.request.Request(
+    os.environ["STATE_HUB_URL"].rstrip("/") + "/progress/",
+    data=body,
+    headers={"Content-Type": "application/json"},
+    method="POST",
+)
+with urllib.request.urlopen(req, timeout=20) as resp:
+    sys.stdout.write(resp.read().decode("utf-8"))
+PY
+}
+
+if [[ -z "$LLM_CONNECT_REMOTE_REPO" ]]; then
+  if [[ -n "$CLUSTER_HOST" ]]; then
+    LLM_CONNECT_REMOTE_REPO="$(ssh "$CLUSTER_HOST" pwd)/llm-connect"
+  else
+    LLM_CONNECT_REMOTE_REPO="$LLM_CONNECT_REPO"
+  fi
+fi
+export LLM_CONNECT_REMOTE_REPO
+
+log "using cluster executor: ${CLUSTER_HOST:-local kubectl}"
+cluster_bash 'set -euo pipefail; command -v kubectl >/dev/null'
+
+log "reconciling non-secret activity-core runtime config"
+cluster_bash "$(cat <<EOF
+set -euo pipefail
+kubectl -n $(quote "$NAMESPACE") patch configmap actcore-runtime-config --type merge -p $(quote "$PATCH_JSON")
+EOF
+)"
+
+LIVE_URL="$(
+  cluster_bash "$(cat <<EOF
+set -euo pipefail
+kubectl -n $(quote "$NAMESPACE") get cm actcore-runtime-config -o jsonpath='{.data.LLM_CONNECT_URL}'
+EOF
+)"
+)"
+LIVE_TIMEOUT="$(
+  cluster_bash "$(cat <<EOF
+set -euo pipefail
+kubectl -n $(quote "$NAMESPACE") get cm actcore-runtime-config -o jsonpath='{.data.LLM_CONNECT_TIMEOUT_SECONDS}'
+EOF
+)"
+)"
+export LIVE_URL LIVE_TIMEOUT
+
+if [[ "$LIVE_URL" != "$EXPECTED_URL" || "$LIVE_TIMEOUT" != "$EXPECTED_TIMEOUT" ]]; then
+  EVIDENCE_STATUS="failed"
+  FAILING_GATE="runtime config mismatch"
+  export EVIDENCE_STATUS FAILING_GATE
+  post_evidence
+  exit 1
+fi
+
+SECRET_KEY_COUNT="$(
+  cluster_bash "$(cat <<EOF
+set -euo pipefail
+if kubectl -n $(quote "$NAMESPACE") get secret $(quote "$SECRET_NAME") >/dev/null 2>&1; then
+  kubectl -n $(quote "$NAMESPACE") get secret $(quote "$SECRET_NAME") -o go-template='{{ len .data }}'
+else
+  printf missing
+fi
+EOF
+)"
+)"
+if [[ "$SECRET_KEY_COUNT" == "missing" ]]; then
+  SECRET_STATUS="missing"
+  SECRET_KEY_COUNT="0"
+elif [[ "${SECRET_KEY_COUNT:-0}" == "0" ]]; then
+  SECRET_STATUS="empty"
+else
+  SECRET_STATUS="present"
+fi
+export SECRET_STATUS SECRET_KEY_COUNT
+
+if [[ "$SECRET_STATUS" != "present" ]]; then
+  EVIDENCE_STATUS="blocked"
+  FAILING_GATE="provider Secret ${SECRET_NAME} ${SECRET_STATUS}"
+  DEPLOYMENT_STATUS="not checked; provider Secret gate not satisfied"
+  SMOKE_STATUS="blocked"
+  SMOKE_SUMMARY="provider Secret must be populated outside Git/State Hub before deployment and smoke"
+  export EVIDENCE_STATUS FAILING_GATE DEPLOYMENT_STATUS SMOKE_STATUS SMOKE_SUMMARY
+  post_evidence
+  [[ "$REQUIRE_SMOKE" == "1" ]] && exit 1
+  exit 0
+fi
+
+if [[ "$APPLY_LLM_CONNECT_OVERLAY" == "1" ]]; then
+  if [[ -n "$CLUSTER_HOST" ]]; then
+    log "syncing llm-connect overlay to ${CLUSTER_HOST}:${LLM_CONNECT_REMOTE_REPO}/deploy/k8s/activity-core-llm-connect"
+    ssh "$CLUSTER_HOST" "mkdir -p $(quote "$LLM_CONNECT_REMOTE_REPO")/deploy/k8s/activity-core-llm-connect"
+    rsync -a --delete \
+      "$LLM_CONNECT_REPO/deploy/k8s/activity-core-llm-connect/" \
+      "${CLUSTER_HOST}:${LLM_CONNECT_REMOTE_REPO}/deploy/k8s/activity-core-llm-connect/"
+  fi
+  log "applying llm-connect overlay"
+  cluster_bash "$(cat <<EOF
+set -euo pipefail
+kubectl apply -k $(quote "$LLM_CONNECT_REMOTE_REPO")/deploy/k8s/activity-core-llm-connect
+kubectl -n $(quote "$NAMESPACE") rollout status deploy/$(quote "$DEPLOYMENT_NAME") --timeout=180s
+EOF
+)"
+fi
+
+DEPLOYMENT_STATUS="$(
+  cluster_bash "$(cat <<EOF
+set -euo pipefail
+if kubectl -n $(quote "$NAMESPACE") get deploy $(quote "$DEPLOYMENT_NAME") >/dev/null 2>&1; then
+  kubectl -n $(quote "$NAMESPACE") get deploy $(quote "$DEPLOYMENT_NAME") -o jsonpath='{.status.readyReplicas}/{.status.replicas}'
+else
+  printf missing
+fi
+EOF
+)"
+)"
+export DEPLOYMENT_STATUS
+
+if [[ "$DEPLOYMENT_STATUS" == "missing" || "$DEPLOYMENT_STATUS" != "1/1" ]]; then
+  EVIDENCE_STATUS="blocked"
+  FAILING_GATE="llm-connect deployment not ready (${DEPLOYMENT_STATUS})"
+  SMOKE_STATUS="blocked"
+  SMOKE_SUMMARY="deployment must be ready before smoke"
+  export EVIDENCE_STATUS FAILING_GATE SMOKE_STATUS SMOKE_SUMMARY
+  post_evidence
+  [[ "$REQUIRE_SMOKE" == "1" ]] && exit 1
+  exit 0
+fi
+
+log "running in-namespace llm-connect fixture smoke"
+set +e
+SMOKE_OUTPUT="$(
+  cluster_bash "$(cat <<EOF
+set -euo pipefail
+kubectl -n $(quote "$NAMESPACE") run llm-connect-smoke-\$(date +%s) \\
+  --rm -i --restart=Never \\
+  --image=llm-connect:latest \\
+  --image-pull-policy=Never \\
+  --env=LLM_CONNECT_URL=$(quote "$EXPECTED_URL") \\
+  --env=LLM_CONNECT_TIMEOUT_SECONDS=$(quote "$EXPECTED_TIMEOUT") \\
+  -- python scripts/smoke_activity_core_endpoint.py
+EOF
+)" 2>&1
+)"
+SMOKE_CODE=$?
+set -e
+
+if [[ "$SMOKE_CODE" == "0" ]]; then
+  SMOKE_STATUS="passed"
+  SMOKE_SUMMARY="$SMOKE_OUTPUT"
+  EVIDENCE_STATUS="passed"
+  FAILING_GATE=""
+else
+  SMOKE_STATUS="failed"
+  SMOKE_SUMMARY="$(printf '%s' "$SMOKE_OUTPUT" | tail -n 5)"
+  EVIDENCE_STATUS="failed"
+  FAILING_GATE="llm-connect fixture smoke failed"
+fi
+export SMOKE_STATUS SMOKE_SUMMARY EVIDENCE_STATUS FAILING_GATE
+post_evidence
+exit "$SMOKE_CODE"
+
--- a/workplans/RAILIANCE-WP-0014-activity-core-llm-connect-live-reconcile.md
+++ b/workplans/RAILIANCE-WP-0014-activity-core-llm-connect-live-reconcile.md
@@ -0,0 +1,96 @@
+---
+id: RAILIANCE-WP-0014
+type: workplan
+title: "activity-core llm-connect live reconcile"
+domain: railiance
+repo: railiance-cluster
+status: blocked
+owner: codex
+topic_slug: railiance
+created: "2026-06-18"
+updated: "2026-06-18"
+state_hub_workstream_id: "a152ddda-d60a-4a65-9b9c-59e2db9ff2b7"
+---
+
+# activity-core llm-connect live reconcile
+
+## Context
+
+activity-core has updated its Railiance runtime manifest so
+`actcore-runtime-config` points at the verified in-cluster llm-connect URL:
+
+```text
+LLM_CONNECT_URL=http://llm-connect.activity-core.svc.cluster.local:8080
+LLM_CONNECT_TIMEOUT_SECONDS=300
+```
+
+The remaining live gate belongs at the cluster/operator layer. Provider
+credentials must stay outside Git and State Hub, and the fixture smoke should
+record only non-secret evidence.
+
+## Add cluster-owned reconcile/check command
+
+```task
+id: RAILIANCE-WP-0014-T01
+status: done
+priority: high
+state_hub_task_id: "49288db7-8102-4ad5-af08-1fe6ab3f1d37"
+```
+
+Add a repeatable Railiance command that:
+
+- reconciles the non-secret activity-core runtime config keys;
+- checks the provider Secret by key count only;
+- applies the llm-connect overlay only after the provider Secret exists;
+- runs the in-namespace fixture smoke only after deployment readiness;
+- posts a non-secret State Hub evidence note.
+
+2026-06-18: Added `tools/cmd/railiance-reconcile-activity-core-llm-connect`
+and Makefile target `reconcile-activity-core-llm-connect`.
+
+## Reconcile live non-secret runtime config
+
+```task
+id: RAILIANCE-WP-0014-T02
+status: done
+priority: high
+state_hub_task_id: "61df5bad-535f-4ad1-ac7a-f46ff278c388"
+```
+
+Patch the live `activity-core/actcore-runtime-config` ConfigMap so it consumes
+the verified llm-connect service URL and timeout. Do not touch Secret values.
+
+2026-06-18: The reconcile command patches only `LLM_CONNECT_URL` and
+`LLM_CONNECT_TIMEOUT_SECONDS`, then re-reads the live ConfigMap to verify the
+values. Live evidence note `c72c514a-399e-4c54-8d5b-d36405932360` confirms
+`LLM_CONNECT_URL=http://llm-connect.activity-core.svc.cluster.local:8080` and
+`LLM_CONNECT_TIMEOUT_SECONDS=300`.
+
+## Complete provider Secret, deployment, and smoke gate
+
+```task
+id: RAILIANCE-WP-0014-T03
+status: blocked
+priority: high
+state_hub_task_id: "ae8af00a-c14f-4b76-933c-46d06cd360ae"
+```
+
+After an operator stores provider credentials in
+`activity-core/llm-connect-provider-secrets`, rerun:
+
+```bash
+make reconcile-activity-core-llm-connect
+```
+
+The command will apply the llm-connect overlay, wait for deployment readiness,
+run the in-namespace fixture smoke with `imagePullPolicy=Never`, and post
+non-secret evidence: provider Secret key count, deployment readiness,
+pass/fail, latency/recommendation summary or sanitized failure.
+
+Current live gate on 2026-06-18: provider Secret
+`activity-core/llm-connect-provider-secrets` is missing, so deployment and
+smoke are intentionally blocked until operator/OpenBao-to-Kubernetes Secret
+custody is complete. Evidence note
+`c72c514a-399e-4c54-8d5b-d36405932360` records provider Secret status
+`missing`, key count `0`, deployment status `not checked; provider Secret gate
+not satisfied`, and smoke status `blocked`.