Documentation
¶
Index ¶
- Variables
- type AlertEvent
- type AlertFilter
- type AlertStore
- type Evaluator
- type GatewayProvider
- type IncidentCreator
- type MetricsProvider
- type Monitor
- type MonitorFilter
- type MonitorStatus
- type MonitorStore
- type MonitorType
- type Service
- func (s *Service) Alerts() AlertStore
- func (s *Service) Evaluator() *Evaluator
- func (s *Service) Monitors() MonitorStore
- func (s *Service) SetIncidentCreator(ic IncidentCreator)
- func (s *Service) SetWebhookDispatcher(wd WebhookDispatcher)
- func (s *Service) Start(ctx context.Context)
- func (s *Service) Stop()
- type ServiceSnapshot
- type WebhookDispatcher
Constants ¶
This section is empty.
Variables ¶
var ErrNotFound = errors.New("monitor: not found")
ErrNotFound is returned when a monitor or alert does not exist.
Functions ¶
This section is empty.
Types ¶
type AlertEvent ¶
type AlertEvent struct {
ID string `json:"id"`
MonitorID string `json:"monitor_id"`
MonitorName string `json:"monitor_name"`
Status MonitorStatus `json:"status"`
Value float64 `json:"value"`
Threshold float64 `json:"threshold"`
Service string `json:"service"`
Action string `json:"action,omitempty"`
Message string `json:"message,omitempty"`
CreatedAt time.Time `json:"created_at"`
}
AlertEvent records a single alert transition (trigger or recovery).
type AlertFilter ¶
type AlertFilter struct {
MonitorID string
Status MonitorStatus
Service string
Limit int
}
AlertFilter controls which alert events are returned by ListAlerts.
type AlertStore ¶
type AlertStore interface {
SaveAlert(ctx context.Context, a *AlertEvent) error
GetAlert(ctx context.Context, id string) (*AlertEvent, error)
ListAlerts(ctx context.Context, filter AlertFilter) ([]AlertEvent, error)
}
AlertStore persists and retrieves alert events.
type Evaluator ¶
type Evaluator struct {
// contains filtered or unexported fields
}
Evaluator periodically checks all enabled monitors against live metrics, creating alert events when thresholds are breached and resolving them on recovery.
func NewEvaluator ¶
func NewEvaluator( monitors MonitorStore, alerts AlertStore, provider MetricsProvider, interval time.Duration, ) *Evaluator
NewEvaluator creates an evaluator that ticks at the given interval.
func (*Evaluator) SetIncidentCreator ¶
func (e *Evaluator) SetIncidentCreator(ic IncidentCreator)
SetIncidentCreator configures the optional incident integration.
func (*Evaluator) SetWebhookDispatcher ¶
func (e *Evaluator) SetWebhookDispatcher(wd WebhookDispatcher)
SetWebhookDispatcher configures the optional webhook integration.
type GatewayProvider ¶
type GatewayProvider struct {
// contains filtered or unexported fields
}
GatewayProvider reads live metrics from the gateway SLO engine and request stats to produce ServiceSnapshots for the evaluator.
func NewGatewayProvider ¶
func NewGatewayProvider(slo *gateway.SLOEngine, stats *gateway.RequestStats) *GatewayProvider
NewGatewayProvider creates a MetricsProvider backed by gateway components.
func (*GatewayProvider) Snapshot ¶
func (g *GatewayProvider) Snapshot(_ context.Context) ([]ServiceSnapshot, error)
Snapshot collects the current metrics from the SLO engine windows and request stats.
type IncidentCreator ¶
type IncidentCreator interface {
CreateFromMonitor(ctx context.Context, alert AlertEvent) error
}
IncidentCreator is the interface the monitor system uses to create incidents when a critical threshold is breached.
type MetricsProvider ¶
type MetricsProvider interface {
Snapshot(ctx context.Context) ([]ServiceSnapshot, error)
}
MetricsProvider is the interface the evaluator calls each tick to collect current metric values. Implementations may read from the gateway stats, DataPlane, or Prometheus.
type Monitor ¶
type Monitor struct {
ID string `json:"id"`
Name string `json:"name"`
Type MonitorType `json:"type"`
Service string `json:"service"` // target service (e.g. "dynamodb"), "*" for all
Action string `json:"action,omitempty"` // target action (e.g. "Query"), "*" or empty for all
Operator string `json:"operator"` // "gt", "lt", "gte", "lte"
Warning float64 `json:"warning"` // warning threshold
Critical float64 `json:"critical"` // critical threshold
Enabled bool `json:"enabled"`
Status MonitorStatus `json:"status"`
LastValue float64 `json:"last_value"`
MutedUntil *time.Time `json:"muted_until,omitempty"`
Tags []string `json:"tags,omitempty"`
Message string `json:"message,omitempty"` // optional description / runbook
CreatedAt time.Time `json:"created_at"`
UpdatedAt time.Time `json:"updated_at"`
LastChecked *time.Time `json:"last_checked,omitempty"`
}
Monitor defines a threshold-based alert rule over a service metric.
type MonitorFilter ¶
type MonitorFilter struct {
Service string
Type MonitorType
Status MonitorStatus
Enabled *bool
Limit int
}
MonitorFilter controls which monitors are returned by List.
type MonitorStatus ¶
type MonitorStatus string
MonitorStatus represents the current evaluation state of a monitor.
const ( MonitorStatusOK MonitorStatus = "ok" MonitorStatusWarning MonitorStatus = "warning" MonitorStatusCritical MonitorStatus = "critical" MonitorStatusMuted MonitorStatus = "muted" MonitorStatusNoData MonitorStatus = "no_data" )
type MonitorStore ¶
type MonitorStore interface {
Save(ctx context.Context, m *Monitor) error
Get(ctx context.Context, id string) (*Monitor, error)
List(ctx context.Context, filter MonitorFilter) ([]Monitor, error)
Update(ctx context.Context, m *Monitor) error
Delete(ctx context.Context, id string) error
ListEnabled(ctx context.Context) ([]Monitor, error)
}
MonitorStore persists and retrieves monitors.
type MonitorType ¶
type MonitorType string
MonitorType identifies the kind of metric a monitor watches.
const ( MonitorTypeErrorRate MonitorType = "error_rate" MonitorTypeLatencyP50 MonitorType = "latency_p50" MonitorTypeLatencyP95 MonitorType = "latency_p95" MonitorTypeLatencyP99 MonitorType = "latency_p99" MonitorTypeThroughput MonitorType = "throughput" MonitorTypeCustom MonitorType = "custom" )
type Service ¶
type Service struct {
// contains filtered or unexported fields
}
Service owns the monitor and alert stores, the evaluator loop, and provides the high-level API used by the admin handlers.
func NewService ¶
func NewService( monitors MonitorStore, alerts AlertStore, provider MetricsProvider, interval time.Duration, ) *Service
NewService creates a monitoring service. Pass nil for provider to skip automatic evaluation (useful for testing or when metrics are not available).
func (*Service) Alerts ¶
func (s *Service) Alerts() AlertStore
Alerts returns the underlying alert store.
func (*Service) Evaluator ¶
Evaluator returns the evaluator, which may be nil if no provider was given.
func (*Service) Monitors ¶
func (s *Service) Monitors() MonitorStore
Monitors returns the underlying monitor store.
func (*Service) SetIncidentCreator ¶
func (s *Service) SetIncidentCreator(ic IncidentCreator)
SetIncidentCreator wires the incident integration into the evaluator.
func (*Service) SetWebhookDispatcher ¶
func (s *Service) SetWebhookDispatcher(wd WebhookDispatcher)
SetWebhookDispatcher wires the webhook integration into the evaluator.
type ServiceSnapshot ¶
type ServiceSnapshot struct {
Service string
ErrorRate float64 // 0..1
LatencyP50 float64 // milliseconds
LatencyP95 float64 // milliseconds
LatencyP99 float64 // milliseconds
Throughput float64 // requests per second
}
ServiceSnapshot holds the current metric values for a single service, collected by the MetricsProvider during each evaluation tick.
Source Files
Directories
¶
| Path | Synopsis |
|---|---|
|
Package dynamostore implements monitor.MonitorStore and monitor.AlertStore backed by DynamoDB via the generic dynamostore package.
|
Package dynamostore implements monitor.MonitorStore and monitor.AlertStore backed by DynamoDB via the generic dynamostore package. |
|
Package filestore implements monitor.MonitorStore and monitor.AlertStore backed by JSON files on disk via the generic filestore package.
|
Package filestore implements monitor.MonitorStore and monitor.AlertStore backed by JSON files on disk via the generic filestore package. |