@oneuptime/common 11.0.1 → 11.0.3

package/Types/Monitor/ProxmoxMetricCatalog.ts

@@ -0,0 +1,382 @@
+import { ProxmoxResourceScope } from "./MonitorStepProxmoxMonitor";
+import MetricsAggregationType from "../Metrics/MetricsAggregationType";
+
+export type ProxmoxMetricCategory =
+  | "Availability"
+  | "Node"
+  | "Guest"
+  | "Storage"
+  | "HA"
+  | "Backup"
+  | "Replication";
+
+export interface ProxmoxMetricDefinition {
+  id: string;
+  friendlyName: string;
+  description: string;
+  metricName: string;
+  category: ProxmoxMetricCategory;
+  defaultAggregation: MetricsAggregationType;
+  defaultResourceScope: ProxmoxResourceScope;
+  unit?: string;
+}
+
+/*
+ * Metric names follow the prometheus-pve-exporter naming scheme. Each series
+ * carries an `id` datapoint label identifying the Proxmox resource it belongs
+ * to: `node/<name>` for nodes, `qemu/<vmid>` for VMs, `lxc/<vmid>` for
+ * containers, and `storage/<storage>` for storage volumes. The agent's OTTL
+ * transform additionally stamps `pve.scope` (node | guest | storage |
+ * cluster), `pve.type` (node | qemu | lxc | storage) and `pve.id` (the part
+ * after the slash) as datapoint attributes — `defaultResourceScope` is the
+ * `pve.scope` value the metric is usually filtered to (Cluster = spans
+ * multiple scopes; don't pre-filter).
+ */
+const proxmoxMetricCatalog: Array<ProxmoxMetricDefinition> = [
+  // Availability Metrics
+  {
+    id: "pve-up",
+    friendlyName: "Resource Up",
+    description:
+      "Whether a Proxmox resource is up (1) or down/stopped (0). Reported for nodes and guests — filter on pve.scope (node/guest) or the id label to scope.",
+    metricName: "pve_up",
+    category: "Availability",
+    defaultAggregation: MetricsAggregationType.Min,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+  },
+  {
+    id: "pve-uptime-seconds",
+    friendlyName: "Uptime",
+    description:
+      "How long a node or guest has been running. Drops to zero when the resource is stopped or rebooted.",
+    metricName: "pve_uptime_seconds",
+    category: "Availability",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "seconds",
+  },
+  {
+    id: "pve-version-info",
+    friendlyName: "PVE Version Info",
+    description:
+      "Metadata series carrying the Proxmox VE release in its version/release labels (value is always 1).",
+    metricName: "pve_version_info",
+    category: "Availability",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "count",
+  },
+
+  // Node Metrics
+  {
+    id: "pve-cpu-usage-ratio",
+    friendlyName: "CPU Usage Ratio",
+    description:
+      "CPU usage as a ratio of available CPU (0 to 1, where 1 = all cores fully used). Reported for nodes and guests — filter on pve.scope to pick one.",
+    metricName: "pve_cpu_usage_ratio",
+    category: "Node",
+    defaultAggregation: MetricsAggregationType.Avg,
+    defaultResourceScope: ProxmoxResourceScope.Node,
+    unit: "ratio",
+  },
+  {
+    id: "pve-cpu-usage-limit",
+    friendlyName: "CPU Limit (Cores)",
+    description:
+      "Maximum CPU available to a node or guest, in cores. For guests this is the number of allocated vCPUs.",
+    metricName: "pve_cpu_usage_limit",
+    category: "Node",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Node,
+    unit: "cores",
+  },
+  {
+    id: "pve-memory-usage-bytes",
+    friendlyName: "Memory Usage",
+    description:
+      "Memory currently in use by a node or guest. Compare against Memory Size to gauge memory pressure.",
+    metricName: "pve_memory_usage_bytes",
+    category: "Node",
+    defaultAggregation: MetricsAggregationType.Avg,
+    defaultResourceScope: ProxmoxResourceScope.Node,
+    unit: "bytes",
+  },
+  {
+    id: "pve-memory-size-bytes",
+    friendlyName: "Memory Size",
+    description:
+      "Total memory available to a node or guest. For guests this is the configured memory allocation.",
+    metricName: "pve_memory_size_bytes",
+    category: "Node",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Node,
+    unit: "bytes",
+  },
+  {
+    id: "pve-node-info",
+    friendlyName: "Node Info",
+    description:
+      "Metadata series for each Proxmox node (value is always 1) carrying the node's name label. Sum it to count the nodes reporting in the cluster.",
+    metricName: "pve_node_info",
+    category: "Node",
+    defaultAggregation: MetricsAggregationType.Sum,
+    defaultResourceScope: ProxmoxResourceScope.Node,
+    unit: "count",
+  },
+
+  // Guest Metrics
+  {
+    id: "pve-network-receive-bytes",
+    friendlyName: "Network Received",
+    description:
+      "Cumulative bytes received over the network by a guest (VM or container). Chart as a rate to see throughput.",
+    metricName: "pve_network_receive_bytes",
+    category: "Guest",
+    defaultAggregation: MetricsAggregationType.Sum,
+    defaultResourceScope: ProxmoxResourceScope.Guest,
+    unit: "bytes",
+  },
+  {
+    id: "pve-network-transmit-bytes",
+    friendlyName: "Network Transmitted",
+    description:
+      "Cumulative bytes transmitted over the network by a guest (VM or container). Chart as a rate to see throughput.",
+    metricName: "pve_network_transmit_bytes",
+    category: "Guest",
+    defaultAggregation: MetricsAggregationType.Sum,
+    defaultResourceScope: ProxmoxResourceScope.Guest,
+    unit: "bytes",
+  },
+  {
+    id: "pve-disk-read-bytes",
+    friendlyName: "Disk Read",
+    description:
+      "Cumulative bytes read from disk by a guest (VM or container). Chart as a rate to see read throughput.",
+    metricName: "pve_disk_read_bytes",
+    category: "Guest",
+    defaultAggregation: MetricsAggregationType.Sum,
+    defaultResourceScope: ProxmoxResourceScope.Guest,
+    unit: "bytes",
+  },
+  {
+    id: "pve-disk-write-bytes",
+    friendlyName: "Disk Write",
+    description:
+      "Cumulative bytes written to disk by a guest (VM or container). Chart as a rate to see write throughput.",
+    metricName: "pve_disk_write_bytes",
+    category: "Guest",
+    defaultAggregation: MetricsAggregationType.Sum,
+    defaultResourceScope: ProxmoxResourceScope.Guest,
+    unit: "bytes",
+  },
+  {
+    id: "pve-guest-info",
+    friendlyName: "Guest Info",
+    description:
+      "Metadata series for each guest (value is always 1) carrying name, node, and type labels. Sum it to count guests in the cluster.",
+    metricName: "pve_guest_info",
+    category: "Guest",
+    defaultAggregation: MetricsAggregationType.Sum,
+    defaultResourceScope: ProxmoxResourceScope.Guest,
+    unit: "count",
+  },
+  {
+    id: "pve-onboot-status",
+    friendlyName: "Start-on-Boot Status",
+    description:
+      "Whether a guest is configured to start on node boot (1) or not (0). A stopped guest with onboot=1 is usually unintended downtime.",
+    metricName: "pve_onboot_status",
+    category: "Guest",
+    defaultAggregation: MetricsAggregationType.Min,
+    defaultResourceScope: ProxmoxResourceScope.Guest,
+    unit: "count",
+  },
+
+  // Storage Metrics
+  {
+    id: "pve-disk-usage-bytes",
+    friendlyName: "Disk Usage",
+    description:
+      "Disk space currently used. Reported for storage volumes, node root filesystems, and guest root disks — filter on pve.scope to pick one. For QEMU guests this reads 0 unless the QEMU guest agent is installed.",
+    metricName: "pve_disk_usage_bytes",
+    category: "Storage",
+    defaultAggregation: MetricsAggregationType.Avg,
+    defaultResourceScope: ProxmoxResourceScope.Storage,
+    unit: "bytes",
+  },
+  {
+    id: "pve-disk-size-bytes",
+    friendlyName: "Disk Size",
+    description:
+      "Total disk space available on a storage volume, node root filesystem, or guest root disk.",
+    metricName: "pve_disk_size_bytes",
+    category: "Storage",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Storage,
+    unit: "bytes",
+  },
+  {
+    id: "pve-storage-info",
+    friendlyName: "Storage Info",
+    description:
+      "Metadata series for each storage volume (value is always 1) carrying node and storage labels. Sum it to count storage volumes in the cluster.",
+    metricName: "pve_storage_info",
+    category: "Storage",
+    defaultAggregation: MetricsAggregationType.Sum,
+    defaultResourceScope: ProxmoxResourceScope.Storage,
+    unit: "count",
+  },
+
+  // HA Metrics
+  {
+    id: "pve-ha-state",
+    friendlyName: "HA State",
+    description:
+      "High-availability state of a node or guest as an enum-style series: one series per possible state (e.g. started, stopped, error) with value 1 for the current state. Filter on the state label to alert on specific states.",
+    metricName: "pve_ha_state",
+    category: "HA",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+  },
+
+  /*
+   * Backup Metrics — cluster-level backup-info collector (default-on).
+   * These report backup-JOB coverage only: whether backups ran
+   * recently or succeeded is not exposed by pve-exporter.
+   */
+  {
+    id: "pve-not-backed-up-total",
+    friendlyName: "Guests Without Backup",
+    description:
+      "Count of guests (VMs and containers) not covered by ANY backup job. A single cluster-level series with no id label. Covers backup-job membership only — not whether backups ran recently or succeeded.",
+    metricName: "pve_not_backed_up_total",
+    category: "Backup",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "count",
+  },
+  {
+    id: "pve-not-backed-up-info",
+    friendlyName: "Guest Without Backup Info",
+    description:
+      "Metadata series present for each guest NOT covered by any backup job (value is always 1), labeled only with the guest's id. Group by id to list the uncovered guests; the series disappears once the guest joins a backup job.",
+    metricName: "pve_not_backed_up_info",
+    category: "Backup",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "count",
+  },
+
+  /*
+   * Replication Metrics — node-level replication collector
+   * (default-on). Series carry the replication JOB id in the `id`
+   * label (e.g. 100-0), not a node/qemu/lxc-prefixed resource id;
+   * pve_replication_info additionally carries type, source, target
+   * and guest labels.
+   */
+  {
+    id: "pve-replication-failed-syncs",
+    friendlyName: "Replication Failed Syncs",
+    description:
+      "Number of consecutive failed sync attempts for a storage replication job. Anything above 0 means the job's replica is going stale — group by id to alert per job.",
+    metricName: "pve_replication_failed_syncs",
+    category: "Replication",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "count",
+  },
+  {
+    id: "pve-replication-duration-seconds",
+    friendlyName: "Replication Duration",
+    description:
+      "How long a replication job's last sync took. A rising duration usually means growing deltas or a slow/congested target.",
+    metricName: "pve_replication_duration_seconds",
+    category: "Replication",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "seconds",
+  },
+  {
+    id: "pve-replication-last-sync-timestamp-seconds",
+    friendlyName: "Replication Last Sync",
+    description:
+      "Unix timestamp of a replication job's last SUCCESSFUL sync. Compare against the current time to gauge replica staleness (the criteria engine has no wall-clock math, so staleness is a chart/UI concern, not an alert).",
+    metricName: "pve_replication_last_sync_timestamp_seconds",
+    category: "Replication",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "seconds",
+  },
+  {
+    id: "pve-replication-last-try-timestamp-seconds",
+    friendlyName: "Replication Last Try",
+    description:
+      "Unix timestamp of a replication job's last sync ATTEMPT. A last-try newer than last-sync means the most recent attempt failed.",
+    metricName: "pve_replication_last_try_timestamp_seconds",
+    category: "Replication",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "seconds",
+  },
+  {
+    id: "pve-replication-next-sync-timestamp-seconds",
+    friendlyName: "Replication Next Sync",
+    description: "Unix timestamp of a replication job's next scheduled sync.",
+    metricName: "pve_replication_next_sync_timestamp_seconds",
+    category: "Replication",
+    defaultAggregation: MetricsAggregationType.Max,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "seconds",
+  },
+  {
+    id: "pve-replication-info",
+    friendlyName: "Replication Info",
+    description:
+      "Metadata series for each storage replication job (value is always 1) carrying type, source, target and guest labels. Sum it to count replication jobs; join its labels to map a job id to the guest and node pair it replicates.",
+    metricName: "pve_replication_info",
+    category: "Replication",
+    defaultAggregation: MetricsAggregationType.Sum,
+    defaultResourceScope: ProxmoxResourceScope.Cluster,
+    unit: "count",
+  },
+];
+
+export function getAllProxmoxMetrics(): Array<ProxmoxMetricDefinition> {
+  return proxmoxMetricCatalog;
+}
+
+export function getProxmoxMetricsByCategory(
+  category: ProxmoxMetricCategory,
+): Array<ProxmoxMetricDefinition> {
+  return proxmoxMetricCatalog.filter((m: ProxmoxMetricDefinition) => {
+    return m.category === category;
+  });
+}
+
+export function getProxmoxMetricById(
+  id: string,
+): ProxmoxMetricDefinition | undefined {
+  return proxmoxMetricCatalog.find((m: ProxmoxMetricDefinition) => {
+    return m.id === id;
+  });
+}
+
+export function getProxmoxMetricByMetricName(
+  metricName: string,
+): ProxmoxMetricDefinition | undefined {
+  return proxmoxMetricCatalog.find((m: ProxmoxMetricDefinition) => {
+    return m.metricName === metricName;
+  });
+}
+
+export function getAllProxmoxMetricCategories(): Array<ProxmoxMetricCategory> {
+  return [
+    "Availability",
+    "Node",
+    "Guest",
+    "Storage",
+    "HA",
+    "Backup",
+    "Replication",
+  ];
+}