Add a VictoriaLogs cluster and point logs-ingest at it (#488)

The k8s log pipeline stores to ClickHouse via NATS+vector, while the VM estate ships journald to a separate puppet-managed VictoriaLogs cluster. Consolidating on VictoriaLogs in-cluster collapses the two paths, and the logs-ingest gateway has no clients yet so it can be repointed now, ahead of the puppet change.

- add VLCluster `logs` at v1.52.0 (2 vlinsert, 2 vlselect, 3 vlstorage, 180d retention, 250Gi each on cephrbd-fast-delete)
- cap vlstorage disk use at 220GiB per node so 180d stays time-based rather than disk-bound
- repoint the logs-ingest HTTPRoute at `vlinsert-logs:9481`
- add a VictoriaLogs Grafana datasource and install its plugin

Nothing is removed here; NATS, ClickHouse, vector, logarchiver and logviewer keep running until a follow-up drops them.

Reviewed-on: #488
Co-authored-by: unkin-agent <unkin-agent@unkin.net>
Co-committed-by: unkin-agent <unkin-agent@unkin.net>

apps/base/grafana/grafanadatasource.yaml

@@ -20,3 +20,22 @@ spec:
     jsonData:
       timeInterval: "15s"
       httpMethod: "POST"
+---
+apiVersion: grafana.integreatly.org/v1beta1
+kind: GrafanaDatasource
+metadata:
+  name: victorialogs
+  namespace: grafana
+spec:
+  instanceSelector:
+    matchLabels:
+      dashboards: "grafana"
+  plugins:
+    - name: victoriametrics-logs-datasource
+      version: 0.32.0
+  datasource:
+    name: "VictoriaLogs"
+    type: "victoriametrics-logs-datasource"
+    uid: "victorialogs"
+    access: "proxy"
+    url: "http://vlselect-logs.logging.svc.cluster.local:9471"

apps/base/logging/gateway.yaml

@@ -1,16 +1,13 @@
 ---
-# Log ingestion endpoint for puppet-managed VMs (and any non-k8s client).
-# Reuses the internal Traefik gateway + cert-manager + external-dns pattern so
-# VMs reach the Vector aggregator's HTTP source over TLS at a DNS name they can
-# resolve. The puppet-side Vector rollout ships NDJSON to
-# https://logs-ingest.k8s.syd1.au.unkin.net/ (a later task).
+# Log ingestion endpoint for puppet-managed VMs (and any non-k8s client):
+# fronts the VLCluster vlinsert service over TLS at a name VMs can resolve.
 apiVersion: gateway.networking.k8s.io/v1
 kind: Gateway
 metadata:
   name: logs-ingest
   namespace: logging
   labels:
-    app.kubernetes.io/name: vector-aggregator
+    app.kubernetes.io/name: victorialogs
     app.kubernetes.io/component: ingest
     traefik.io/instance: internal
   annotations:

apps/base/logging/httproute.yaml

@@ -5,7 +5,7 @@ metadata:
   name: logs-ingest-http-redirect
   namespace: logging
   labels:
-    app.kubernetes.io/name: vector-aggregator
+    app.kubernetes.io/name: victorialogs
     app.kubernetes.io/component: ingest
 spec:
   hostnames:
@@ -32,7 +32,7 @@ metadata:
   name: logs-ingest
   namespace: logging
   labels:
-    app.kubernetes.io/name: vector-aggregator
+    app.kubernetes.io/name: victorialogs
     app.kubernetes.io/component: ingest
 spec:
   hostnames:
@@ -46,8 +46,8 @@ spec:
     - backendRefs:
         - group: ""
           kind: Service
-          name: vector-vm-ingest
-          port: 8080
+          name: vlinsert-logs
+          port: 9481
           weight: 1
       matches:
         - path:

apps/base/logging/kustomization.yaml

@@ -10,6 +10,7 @@ resources:
   - job_clickhouse-schema.yaml
   - nats-bootstrap-job.yaml
   - cephrgw.yaml
+  - vlcluster.yaml
   - gateway.yaml
   - httproute.yaml
   - serviceaccount_logarchiver.yaml

apps/base/logging/vlcluster.yaml

@@ -0,0 +1,47 @@
+---
+apiVersion: operator.victoriametrics.com/v1
+kind: VLCluster
+metadata:
+  name: logs
+  namespace: logging
+spec:
+  clusterVersion: v1.52.0
+  vlinsert:
+    replicaCount: 2
+    resources:
+      requests:
+        cpu: 500m
+        memory: 1Gi
+      limits:
+        cpu: "2"
+        memory: 4Gi
+  vlselect:
+    replicaCount: 2
+    resources:
+      requests:
+        cpu: 500m
+        memory: 1Gi
+      limits:
+        cpu: "2"
+        memory: 4Gi
+  vlstorage:
+    replicaCount: 3
+    retentionPeriod: 180d
+    # ~3 GiB/day measured; 220GiB/node cap keeps 180d time-based, not disk-bound
+    retentionMaxDiskSpaceUsageBytes: 220GiB
+    storage:
+      volumeClaimTemplate:
+        spec:
+          accessModes:
+            - ReadWriteOnce
+          storageClassName: cephrbd-fast-delete
+          resources:
+            requests:
+              storage: 250Gi
+    resources:
+      requests:
+        cpu: "1"
+        memory: 2Gi
+      limits:
+        cpu: "4"
+        memory: 8Gi