Skip to content

armgit / Pulse

Public Imported from GitHub
0

Real-time monitoring dashboard for Proxmox VE, PBS, Docker, Kubernetes, TrueNAS and vSphere. Self-hosted, with alerts for failed backups, full disks, unhealthy containers and offline nodes.

8.3 KB
Raw Blame
package monitoring

import (
	"strings"
	"time"

	"github.com/rcourtman/pulse-go-rewrite/internal/alerts"
	"github.com/rcourtman/pulse-go-rewrite/internal/logging"
	"github.com/rcourtman/pulse-go-rewrite/internal/models"
	"github.com/rcourtman/pulse-go-rewrite/internal/unifiedresources"
	"github.com/rs/zerolog"
	"github.com/rs/zerolog/log"
)

// syncAlertsToState copies the latest alert manager data into the shared state snapshot.
// This keeps WebSocket broadcasts aligned with in-memory acknowledgement updates.
func (m *Monitor) syncAlertsToState() {
	if m.pruneStaleDockerAlerts() {
		if logging.IsLevelEnabled(zerolog.DebugLevel) {
			log.Debug().Msg("pruned stale docker alerts during sync")
		}
	}

	modelAlerts := m.activeAlertsSnapshot()
	for _, alert := range modelAlerts {
		if alert.Acknowledged && logging.IsLevelEnabled(zerolog.DebugLevel) {
			log.Debug().Str("alertID", alert.ID).Interface("ackTime", alert.AckTime).Msg("syncing acknowledged alert")
		}
	}
	m.state.UpdateActiveAlerts(modelAlerts)

	recentlyResolved := m.alertManager.GetRecentlyResolved()
	if len(recentlyResolved) > 0 && logging.IsLevelEnabled(zerolog.DebugLevel) {
		log.Debug().Int("count", len(recentlyResolved)).Msg("syncing recently resolved alerts")
	}
	m.state.UpdateRecentlyResolved(recentlyResolved)
}

// SyncAlertState is the exported wrapper used by APIs that mutate alerts outside the poll loop.
func (m *Monitor) SyncAlertState() {
	m.syncAlertsToState()
}

func (m *Monitor) activeAlertsSnapshot() []models.Alert {
	if m == nil {
		return nil
	}
	if m.alertManager == nil {
		if m.state == nil {
			return nil
		}
		return m.state.GetSnapshot().ActiveAlerts
	}

	activeAlerts := m.alertManager.GetActiveAlerts()
	modelAlerts := make([]models.Alert, 0, len(activeAlerts))
	for _, alert := range activeAlerts {
		modelAlerts = append(modelAlerts, models.Alert{
			ID:              alert.ID,
			Type:            alert.Type,
			Level:           string(alert.Level),
			ResourceID:      alert.ResourceID,
			ResourceName:    alert.ResourceName,
			Node:            alert.Node,
			NodeDisplayName: alert.NodeDisplayName,
			Instance:        alert.Instance,
			Message:         alert.Message,
			Summary:         alert.Summary,
			Value:           alert.Value,
			Threshold:       alert.Threshold,
			StartTime:       alert.StartTime,
			Acknowledged:    alert.Acknowledged,
			AckTime:         alert.AckTime,
			AckUser:         alert.AckUser,
			// GetActiveAlerts returns deep clones, so the map and the live
			// metric status are already private.
			Metadata:     alert.Metadata,
			MetricStatus: alert.MetricStatus,
		})
	}
	return modelAlerts
}

func (m *Monitor) recentlyResolvedAlertsSnapshot() []models.ResolvedAlert {
	if m == nil {
		return nil
	}
	if m.alertManager == nil {
		if m.state == nil {
			return nil
		}
		return m.state.GetSnapshot().RecentlyResolved
	}

	return m.alertManager.GetRecentlyResolved()
}

// syncUnifiedResourceAlertsToState evaluates resources read under scope. A
// mock-mode flip since that read refuses every remaining step, so a snapshot
// of the mode the monitor left can neither open or resolve alerts nor rewrite
// persisted identities and links in the mode it entered.
func (m *Monitor) syncUnifiedResourceAlertsToState(resources []unifiedresources.Resource, scope mockModeScope) {
	if m == nil || m.alertManager == nil {
		return
	}

	// Alert intent resolves through the published registry with saved hosts
	// overlaid, under the alert manager's lock. Build that overlay for the new
	// generation here, outside the lock, so its lookups need not.
	m.operatorIntentIdentity.Load().refresh()

	// Resource publication can make persisted canonical policy resolvable for
	// restored native-ID alerts that preceded the first observation at startup.
	scope.run(func() { m.alertManager.ReconcileOperatorIntentState() })
	scope.run(func() { m.migrateAlertIdentities(resources) })
	scope.run(func() { m.migrateAvailabilityLinksToCanonicalIDs(resources) })

	// Metrics go one resource per fenced call, as
	// CheckUnifiedResourceMetricsWithCapacityTrends would iterate them, so a
	// mode change waits for one resource's evaluation rather than the estate's.
	capacityTrends := m.unifiedStorageCapacityTrends(resources, time.Now())
	for _, resource := range resources {
		input, ok := alerts.UnifiedResourceInputFromResource(resource)
		if !ok {
			continue
		}
		scope.run(func() { m.alertManager.CheckUnifiedResourceWithCapacityTrend(input, capacityTrends[input.ID]) })
	}
	scope.run(func() { m.alertManager.SyncUnifiedResourceIncidents(resources) })
	m.syncAlertsToState()
}

// migrateAlertIdentities persists the alert identity migration that resources
// make resolvable.
func (m *Monitor) migrateAlertIdentities(resources []unifiedresources.Resource) {
	config := m.alertManager.GetConfig()
	migrationPlan := alerts.PlanAlertIdentityMigration(config, resources)
	if migrationPlan.UnsupportedVersion {
		log.Warn().
			Int("configVersion", migrationPlan.FromVersion).
			Int("supportedVersion", alerts.CurrentAlertIdentitySchemaVersion).
			Msg("alert identity config uses a newer schema; leaving persisted identities untouched")
	} else if alerts.ApplyAlertIdentityMigration(&config, migrationPlan) {
		if m.configPersist == nil {
			log.Warn().
				Int("fromVersion", migrationPlan.FromVersion).
				Int("toVersion", migrationPlan.ToVersion).
				Int("removedOverrides", len(migrationPlan.RemovedOverrideKeys)).
				Int("addedOverrides", len(migrationPlan.AddedOverrideKeys)).
				Int("deferredOverrides", len(migrationPlan.Deferred)).
				Msg("cannot persist planned alert identity migration without config persistence")
		} else if err := m.configPersist.SaveAlertConfig(config); err != nil {
			log.Error().
				Err(err).
				Int("fromVersion", migrationPlan.FromVersion).
				Int("toVersion", migrationPlan.ToVersion).
				Msg("failed to persist planned alert identity migration")
		} else {
			m.alertManager.UpdateConfig(config)
			log.Info().
				Int("fromVersion", migrationPlan.FromVersion).
				Int("toVersion", migrationPlan.ToVersion).
				Int("removedOverrides", len(migrationPlan.RemovedOverrideKeys)).
				Int("addedOverrides", len(migrationPlan.AddedOverrideKeys)).
				Int("deferredOverrides", len(migrationPlan.Deferred)).
				Msg("persisted alert identity migration plan")
		}
	}
}

// pruneStaleDockerAlerts removes docker alerts that reference hosts no longer present in state.
func (m *Monitor) pruneStaleDockerAlerts() bool {
	if m.alertManager == nil {
		return false
	}

	// The inventory is the fixture read view in mock mode. A prune that read
	// it must not remove Docker alerts the other mode raised after a switch.
	// Only the hosts' IDs matter, so the view may predate the last metric ticks.
	scope := m.mockModeFence.begin()
	readState := m.currentModeStructureReadState()
	if readState == nil {
		return false
	}

	hosts := readState.DockerHosts()
	knownHosts := make(map[string]struct{}, len(hosts)*2)
	for _, host := range hosts {
		hostID := strings.TrimSpace(host.ID())
		if hostID != "" {
			knownHosts[hostID] = struct{}{}
		}
		if sourceID := strings.TrimSpace(host.HostSourceID()); sourceID != "" {
			knownHosts[sourceID] = struct{}{}
		}
	}

	if len(knownHosts) == 0 {
		// Still allow stale entries to be cleared if no hosts remain.
	}

	active := m.alertManager.GetActiveAlerts()
	processed := make(map[string]struct{})
	cleared := false

	for _, alert := range active {
		var hostID string

		switch {
		case alert.Type == "docker-host-offline":
			hostID = strings.TrimPrefix(strings.TrimSpace(alert.ResourceID), "docker:")
		case strings.HasPrefix(alert.ResourceID, "docker:"):
			resource := strings.TrimPrefix(alert.ResourceID, "docker:")
			if idx := strings.Index(resource, "/"); idx >= 0 {
				hostID = resource[:idx]
			} else {
				hostID = resource
			}
		default:
			continue
		}

		hostID = strings.TrimSpace(hostID)
		if hostID == "" {
			continue
		}

		if _, known := knownHosts[hostID]; known {
			continue
		}
		if _, alreadyCleared := processed[hostID]; alreadyCleared {
			continue
		}

		host := models.DockerHost{
			ID:          hostID,
			DisplayName: alert.ResourceName,
			Hostname:    alert.Node,
		}
		if host.DisplayName == "" {
			host.DisplayName = hostID
		}
		if host.Hostname == "" {
			host.Hostname = hostID
		}

		if !scope.run(func() { m.alertManager.HandleDockerHostRemoved(host) }) {
			return cleared
		}
		processed[hostID] = struct{}{}
		cleared = true
	}

	return cleared
}