Demo rack staging

common/roles/defaults/defaults/main.yaml

@@ -1,10 +1,10 @@
 ---
 metal_registry_auth_enabled: false
 metal_registry_auth_user: ""
 metal_registry_auth_password: ""
 metal_registry_url: "https://index.docker.io/v1/"
 
 metal_stack_release:
   mapping:
     # binaries
     metal_hammer_image_tag: "binaries.metal-stack.metal-hammer.version"
@@ -98,6 +98,8 @@
     capms_controller_name: "docker-images.metal-stack.kubernetes.cluster-api-provider-metal-stack-controller.name"
     capms_controller_tag: "docker-images.metal-stack.kubernetes.cluster-api-provider-metal-stack-controller.tag"
     # third party
+    alloy_image_tag: "docker-images.third-party.partition.alloy.tag"
+    alloy_image_name: "docker-images.third-party.partition.alloy.name"
     event_exporter_name: "docker-images.third-party.control-plane.event-exporter.name"
     event_exporter_tag: "docker-images.third-party.control-plane.event-exporter.tag"
     nsq_image_tag: "docker-images.third-party.control-plane.nsq.tag"

common/roles/partition-prometheus-rules/README.md

@@ -14,3 +14,4 @@ Contains default variables for partition prometheus alert manager rules. This wa
 | partition_prometheus_rules_node_exporter     |           | node-exporter alert rules     |
 | partition_prometheus_rules_pixiecore         |           | pixiecore alert rules         |
 | partition_prometheus_rules_sonic_exporter    |           | sonic-exporter alert rules    |
+| partition_prometheus_rules_goflow2           |           | goflow2 sFlow collector rules |

common/roles/partition-prometheus-rules/defaults/main.yaml

@@ -343,3 +343,39 @@
       annotations:
         description: "Sonic switch {{$labels.instance}} in partition {{$externalLabels.partition}} could not be scraped for more than 10 minutes."
   {% endraw %}
+
+partition_prometheus_rules_sonic_systemd: |
+  {% raw %}
+  - name: sonic-systemd.rules
+    rules:
+    - record: sonic:systemd_service_uptime_seconds
+      expr: time() - node_systemd_unit_start_time_seconds{name=~"(swss|bgp).service"}
+    - alert: SonicServiceNotActive
+      expr: node_systemd_unit_state{name=~"(swss|bgp).service", state="active"} == 0
+      for: 5m
+      labels:
+        severity: critical
+      annotations:
+        description: "SONiC service {{ $labels.name }} on {{ $labels.instance }} is not active."
+    - alert: SonicServiceRestarted
+      expr: changes(node_systemd_unit_start_time_seconds{name=~"(swss|bgp).service"}[8h]) > 0
+      labels:
+        severity: critical
+      annotations:
+        description: "SONiC service {{ $labels.name }} on {{ $labels.instance }} has restarted in the last 8 hours."
+  {% endraw %}
+
+partition_prometheus_rules_goflow2: |
+  {% raw %}
+  - name: goflow2-collector.rules
+    rules:
+    - alert: GoflowCollectorDown
+      expr: up{job="goflow2"} == 0
+      for: 2m
+      labels:
+        severity: critical
+      annotations:
+        summary: "goflow2 sFlow collector down ({{ $labels.instance }})"
+        description: "goflow2 instance {{ $labels.instance }} has been unreachable for more than 2 minutes. sFlow ingestion is halted."
+  {% endraw %}
+

control-plane/README.md

@@ -46,6 +46,7 @@ The `defaults` folder contains defaults that are used by multiple roles in the c
 | [gardener-shoots](roles/gardener-shoots)                               | Deploys managed shoots into the virtual garden.                                                                                    |
 | [gardener-virtual-garden-access](roles/gardener-virtual-garden-access) | Deploys a managed resource to access the the Virtual Garden with operator setup                                                    |
 | [headscale](roles/headscale)                                           | Deploys headscale for firewall VPN                                                                                                 |
+| [lldp-weathermap](roles/lldp-weathermap)                               | Generates a Grafana network topology dashboard from LLDP data with live link utilization                                           |
 | [ipam-db](roles/ipam-db)                                               | Deploys a database for the [IPAM](https://github.com/metal-stack/go-ipam) of the metal-api                                         |
 | [isolated-clusters](roles/isolated-clusters)                           | Deploys services for Gardener isolated clusters                                                                                    |
 | [logging](roles/logging)                                               | Deploys metal-stack control plane logging components                                                                               |

control-plane/roles/lldp-weathermap/README.md

@@ -0,0 +1,19 @@
+# lldp-weathermap
+
+Generates a Grafana canvas dashboard that visualizes network topology and live link utilization. Topology is derived from LLDP data collected from switches at deploy time. Nodes are organized into configurable tiers (e.g. spine, exit, leaf) and links are color-coded against configurable bandwidth thresholds.
+
+You can look up all the default values of this role [here](defaults/main.yaml).
+
+## Variables
+
+| Name                               | Mandatory | Description                                                                                       |
+| ---------------------------------- | --------- | ------------------------------------------------------------------------------------------------- |
+| `lldp_weathermap_namespace`        |           | Kubernetes namespace for the Grafana dashboard ConfigMap                                          |
+| `lldp_weathermap_dashboard_name`   |           | Display name of the Grafana dashboard                                                             |
+| `lldp_weathermap_tiers`            |           | List of tier definitions; each entry has `id`, `top_fraction`, `background_color`, `border_color` |
+| `lldp_weathermap_fallback_tier`    |           | Tier assigned to nodes not matching any known tier id                                             |
+| `lldp_weathermap_panel_grid_rows`  |           | Grid row height of the canvas panel                                                               |
+| `lldp_weathermap_thresholds`       |           | Link utilization thresholds mapping bps values to display colors                                  |
+| `lldp_weathermap_hosts_group`      | yes       | Ansible inventory group to collect LLDP data from                                                 |
+| `lldp_weathermap_node_filter`      |           | List of tier IDs to include in the weathermap; nodes in other tiers are hidden                    |
+| `lldp_weathermap_static_links`     |           | Static link definitions that can't be autodiscovered using LLDP                                   |

control-plane/roles/lldp-weathermap/defaults/main.yaml

@@ -0,0 +1,38 @@
+---
+lldp_weathermap_namespace: monitoring
+lldp_weathermap_dashboard_name: "Network Weathermap"
+
+lldp_weathermap_tiers:
+  - id: spine
+    top_fraction: 0.15
+    background_color: "#1f3a5f"
+    border_color: "#4a9eff"
+  - id: exit
+    top_fraction: 0.45
+    background_color: "#3a1f5f"
+    border_color: "#a855f7"
+  - id: leaf
+    top_fraction: 0.75
+    background_color: "#1f4a4a"
+    border_color: "#2dd4bf"
+
+lldp_weathermap_fallback_tier: exit
+
+lldp_weathermap_panel_grid_rows: 14
+
+lldp_weathermap_thresholds:
+  - color: green
+    value: null
+  - color: yellow
+    value: 9000000000   # 9 Gbps
+  - color: orange
+    value: 18000000000   # 18 Gbps
+  - color: red
+    value: 24000000000   # 24 Gbps
+
+lldp_weathermap_hosts_group: all_switches
+lldp_weathermap_node_filter:
+  - leaf
+  - spine
+
+lldp_weathermap_static_links: []