diff --git a/internal/app.go b/internal/app.go index 7219f55..60bf2da 100644 --- a/internal/app.go +++ b/internal/app.go @@ -83,6 +83,9 @@ func setup(app *orz.App) error { // 启动数据清理任务 go components.AgentService.StartCleanupTask(ctx) + // 启动聚合下采样任务 + go components.AgentService.StartAggregationTask(ctx) + // 启动指标监控任务（用于告警检测） go startMetricsMonitoring(ctx, components, app.Logger()) @@ -274,7 +277,6 @@ func autoMigrate(database *gorm.DB) error { &models.DiskMetric{}, &models.NetworkMetric{}, &models.NetworkConnectionMetric{}, - &models.LoadMetric{}, &models.DiskIOMetric{}, &models.GPUMetric{}, &models.TemperatureMetric{}, @@ -289,6 +291,16 @@ func autoMigrate(database *gorm.DB) error { &models.TamperProtectConfig{}, &models.TamperEvent{}, &models.TamperAlert{}, + // 聚合表 + &models.AggregatedCPUMetricModel{}, + &models.AggregatedMemoryMetricModel{}, + &models.AggregatedDiskMetricModel{}, + &models.AggregatedNetworkMetricModel{}, + &models.AggregatedNetworkConnectionMetricModel{}, + &models.AggregatedDiskIOMetricModel{}, + &models.AggregatedGPUMetricModel{}, + &models.AggregatedTemperatureMetricModel{}, + &models.AggregationProgress{}, ) } diff --git a/internal/handler/agent_handler.go b/internal/handler/agent_handler.go index 8c3b33b..d83a3f7 100644 --- a/internal/handler/agent_handler.go +++ b/internal/handler/agent_handler.go @@ -354,33 +354,8 @@ func (h *AgentHandler) GetForAdmin(c echo.Context) error { return orz.Ok(c, agent) } -// GetMetrics 获取探针聚合指标（公开接口，已登录返回全部，未登录返回公开可见） -func (h *AgentHandler) GetMetrics(c echo.Context) error { - agentID := c.Param("id") - ctx := c.Request().Context() - - // 验证探针访问权限 - if _, err := h.agentService.GetAgentByAuth(ctx, agentID, utils.IsAuthenticated(c)); err != nil { - return err - } - - metricType := c.QueryParam("type") - rangeParam := c.QueryParam("range") - - // 验证指标类型 - validTypes := map[string]bool{ - "cpu": true, "memory": true, "disk": true, "network": true, "network_connection": true, "load": true, - "disk_io": true, "gpu": true, "temperature": true, - } - if metricType == "" { - return orz.NewError(400, "指标类型不能为空") - } - if !validTypes[metricType] { - return orz.NewError(400, "无效的指标类型") - } - - // 根据 range 参数自动计算时间范围 - var start, end int64 +// parseTimeRange 解析时间范围参数，返回起始和结束时间（毫秒） +func parseTimeRange(rangeParam string) (start, end int64, err error) { end = time.Now().UnixMilli() if rangeParam == "" { @@ -398,8 +373,56 @@ func (h *AgentHandler) GetMetrics(c echo.Context) error { start = end - 30*60*1000 case "1h": start = end - 60*60*1000 + case "3h": + start = end - 3*60*60*1000 + case "6h": + start = end - 6*60*60*1000 + case "12h": + start = end - 12*60*60*1000 + case "1d", "24h": + start = end - 24*60*60*1000 + case "3d": + start = end - 3*24*60*60*1000 + case "7d": + start = end - 7*24*60*60*1000 + case "30d": + start = end - 30*24*60*60*1000 default: - return orz.NewError(400, "无效的时间范围，支持: 1m, 5m, 15m, 30m, 1h") + return 0, 0, fmt.Errorf("无效的时间范围，支持: 1m, 5m, 15m, 30m, 1h, 3h, 6h, 12h, 1d/24h, 3d, 7d, 30d") + } + + return start, end, nil +} + +// GetMetrics 获取探针聚合指标（公开接口，已登录返回全部，未登录返回公开可见） +func (h *AgentHandler) GetMetrics(c echo.Context) error { + agentID := c.Param("id") + ctx := c.Request().Context() + + // 验证探针访问权限 + if _, err := h.agentService.GetAgentByAuth(ctx, agentID, utils.IsAuthenticated(c)); err != nil { + return err + } + + metricType := c.QueryParam("type") + rangeParam := c.QueryParam("range") + + // 验证指标类型 + validTypes := map[string]bool{ + "cpu": true, "memory": true, "disk": true, "network": true, "network_connection": true, + "disk_io": true, "gpu": true, "temperature": true, + } + if metricType == "" { + return orz.NewError(400, "指标类型不能为空") + } + if !validTypes[metricType] { + return orz.NewError(400, "无效的指标类型") + } + + // 解析时间范围 + start, end, err := parseTimeRange(rangeParam) + if err != nil { + return orz.NewError(400, err.Error()) } // 服务端自动计算最优聚合间隔 @@ -433,27 +456,10 @@ func (h *AgentHandler) GetNetworkMetricsByInterface(c echo.Context) error { rangeParam := c.QueryParam("range") - // 根据 range 参数自动计算时间范围 - var start, end int64 - end = time.Now().UnixMilli() - - if rangeParam == "" { - rangeParam = "1h" // 默认1小时 - } - - switch rangeParam { - case "1m": - start = end - 1*60*1000 - case "5m": - start = end - 5*60*1000 - case "15m": - start = end - 15*60*1000 - case "30m": - start = end - 30*60*1000 - case "1h": - start = end - 60*60*1000 - default: - return orz.NewError(400, "无效的时间范围，支持: 1m, 5m, 15m, 30m, 1h") + // 解析时间范围 + start, end, err := parseTimeRange(rangeParam) + if err != nil { + return orz.NewError(400, err.Error()) } // 服务端自动计算最优聚合间隔 @@ -701,27 +707,10 @@ func (h *AgentHandler) GetMonitorMetrics(c echo.Context) error { rangeParam := c.QueryParam("range") ctx := c.Request().Context() - // 根据 range 参数自动计算时间范围 - var start, end int64 - end = time.Now().UnixMilli() - - if rangeParam == "" { - rangeParam = "1h" // 默认1小时 - } - - switch rangeParam { - case "1m": - start = end - 1*60*1000 - case "5m": - start = end - 5*60*1000 - case "15m": - start = end - 15*60*1000 - case "30m": - start = end - 30*60*1000 - case "1h": - start = end - 60*60*1000 - default: - return orz.NewError(400, "无效的时间范围，支持: 1m, 5m, 15m, 30m, 1h") + // 解析时间范围 + start, end, err := parseTimeRange(rangeParam) + if err != nil { + return orz.NewError(400, err.Error()) } metrics, err := h.agentService.GetMonitorMetrics(ctx, agentID, monitorName, start, end) diff --git a/internal/models/metric.go b/internal/models/metric.go index c11a2d9..0ca25eb 100644 --- a/internal/models/metric.go +++ b/internal/models/metric.go @@ -88,20 +88,6 @@ func (NetworkConnectionMetric) TableName() string { return "network_connection_metrics" } -// LoadMetric 系统负载指标 -type LoadMetric struct { - ID uint `gorm:"primaryKey;autoIncrement" json:"id"` - AgentID string `gorm:"index:idx_load_agent_ts,priority:1" json:"agentId"` // 探针ID - Load1 float64 `json:"load1"` // 1分钟负载 - Load5 float64 `json:"load5"` // 5分钟负载 - Load15 float64 `json:"load15"` // 15分钟负载 - Timestamp int64 `gorm:"index:idx_load_agent_ts,priority:2;index:idx_load_ts" json:"timestamp"` // 时间戳（毫秒） -} - -func (LoadMetric) TableName() string { - return "load_metrics" -} - // DiskIOMetric 磁盘IO指标 type DiskIOMetric struct { ID uint `gorm:"primaryKey;autoIncrement" json:"id"` @@ -199,3 +185,150 @@ type MonitorMetric struct { func (MonitorMetric) TableName() string { return "monitor_metrics" } + +// ----------- 聚合表 ----------- + +// AggregatedCPUMetricModel CPU聚合表 +type AggregatedCPUMetricModel struct { + ID uint `gorm:"primaryKey;autoIncrement" json:"id"` + AgentID string `gorm:"index:idx_cpuagg_agent_bucket,priority:1;uniqueIndex:ux_cpuagg_bucket,priority:1" json:"agentId"` + BucketSeconds int `gorm:"index:idx_cpuagg_agent_bucket,priority:2;uniqueIndex:ux_cpuagg_bucket,priority:2" json:"bucketSeconds"` + BucketStart int64 `gorm:"index:idx_cpuagg_agent_bucket,priority:3;uniqueIndex:ux_cpuagg_bucket,priority:3" json:"bucketStart"` // 毫秒 + MaxUsage float64 `json:"maxUsage"` + LogicalCores int `json:"logicalCores"` +} + +func (AggregatedCPUMetricModel) TableName() string { + return "cpu_metrics_aggs" +} + +// AggregatedMemoryMetricModel 内存聚合表 +type AggregatedMemoryMetricModel struct { + ID uint `gorm:"primaryKey;autoIncrement" json:"id"` + AgentID string `gorm:"index:idx_memagg_agent_bucket,priority:1;uniqueIndex:ux_memagg_bucket,priority:1" json:"agentId"` + BucketSeconds int `gorm:"index:idx_memagg_agent_bucket,priority:2;uniqueIndex:ux_memagg_bucket,priority:2" json:"bucketSeconds"` + BucketStart int64 `gorm:"index:idx_memagg_agent_bucket,priority:3;uniqueIndex:ux_memagg_bucket,priority:3" json:"bucketStart"` // 毫秒 + MaxUsage float64 `json:"maxUsage"` + Total uint64 `json:"total"` +} + +func (AggregatedMemoryMetricModel) TableName() string { + return "memory_metrics_aggs" +} + +// AggregatedDiskMetricModel 磁盘聚合表 +type AggregatedDiskMetricModel struct { + ID uint `gorm:"primaryKey;autoIncrement" json:"id"` + AgentID string `gorm:"index:idx_diskagg_agent_bucket_mp,priority:1;uniqueIndex:ux_diskagg_bucket,priority:1" json:"agentId"` + BucketSeconds int `gorm:"index:idx_diskagg_agent_bucket_mp,priority:2;uniqueIndex:ux_diskagg_bucket,priority:2" json:"bucketSeconds"` + BucketStart int64 `gorm:"index:idx_diskagg_agent_bucket_mp,priority:3;uniqueIndex:ux_diskagg_bucket,priority:3" json:"bucketStart"` // 毫秒 + MountPoint string `gorm:"index:idx_diskagg_agent_bucket_mp,priority:4;uniqueIndex:ux_diskagg_bucket,priority:4" json:"mountPoint"` + MaxUsage float64 `json:"maxUsage"` + Total uint64 `json:"total"` +} + +func (AggregatedDiskMetricModel) TableName() string { + return "disk_metrics_aggs" +} + +// AggregatedNetworkMetricModel 网络聚合表（按网卡分组） +type AggregatedNetworkMetricModel struct { + ID uint `gorm:"primaryKey;autoIncrement" json:"id"` + AgentID string `gorm:"index:idx_netagg_agent_bucket_iface,priority:1;uniqueIndex:ux_netagg_bucket,priority:1" json:"agentId"` + BucketSeconds int `gorm:"index:idx_netagg_agent_bucket_iface,priority:2;uniqueIndex:ux_netagg_bucket,priority:2" json:"bucketSeconds"` + BucketStart int64 `gorm:"index:idx_netagg_agent_bucket_iface,priority:3;uniqueIndex:ux_netagg_bucket,priority:3" json:"bucketStart"` // 毫秒 + Interface string `gorm:"index:idx_netagg_agent_bucket_iface,priority:4;uniqueIndex:ux_netagg_bucket,priority:4" json:"interface"` + MaxSentRate float64 `json:"maxSentRate"` + MaxRecvRate float64 `json:"maxRecvRate"` +} + +func (AggregatedNetworkMetricModel) TableName() string { + return "network_metrics_aggs" +} + +// AggregatedNetworkConnectionMetricModel 网络连接聚合表 +type AggregatedNetworkConnectionMetricModel struct { + ID uint `gorm:"primaryKey;autoIncrement" json:"id"` + AgentID string `gorm:"index:idx_netconnagg_agent_bucket,priority:1;uniqueIndex:ux_netconnagg_bucket,priority:1" json:"agentId"` + BucketSeconds int `gorm:"index:idx_netconnagg_agent_bucket,priority:2;uniqueIndex:ux_netconnagg_bucket,priority:2" json:"bucketSeconds"` + BucketStart int64 `gorm:"index:idx_netconnagg_agent_bucket,priority:3;uniqueIndex:ux_netconnagg_bucket,priority:3" json:"bucketStart"` // 毫秒 + MaxEstablished uint32 `json:"maxEstablished"` + MaxSynSent uint32 `json:"maxSynSent"` + MaxSynRecv uint32 `json:"maxSynRecv"` + MaxFinWait1 uint32 `json:"maxFinWait1"` + MaxFinWait2 uint32 `json:"maxFinWait2"` + MaxTimeWait uint32 `json:"maxTimeWait"` + MaxClose uint32 `json:"maxClose"` + MaxCloseWait uint32 `json:"maxCloseWait"` + MaxLastAck uint32 `json:"maxLastAck"` + MaxListen uint32 `json:"maxListen"` + MaxClosing uint32 `json:"maxClosing"` + MaxTotal uint32 `json:"maxTotal"` +} + +func (AggregatedNetworkConnectionMetricModel) TableName() string { + return "network_connection_metrics_aggs" +} + +// AggregatedDiskIOMetricModel 磁盘IO聚合表 +type AggregatedDiskIOMetricModel struct { + ID uint `gorm:"primaryKey;autoIncrement" json:"id"` + AgentID string `gorm:"index:idx_diskioagg_agent_bucket_dev,priority:1;uniqueIndex:ux_diskioagg_bucket,priority:1" json:"agentId"` + BucketSeconds int `gorm:"index:idx_diskioagg_agent_bucket_dev,priority:2;uniqueIndex:ux_diskioagg_bucket,priority:2" json:"bucketSeconds"` + BucketStart int64 `gorm:"index:idx_diskioagg_agent_bucket_dev,priority:3;uniqueIndex:ux_diskioagg_bucket,priority:3" json:"bucketStart"` // 毫秒 + Device string `gorm:"index:idx_diskioagg_agent_bucket_dev,priority:4;uniqueIndex:ux_diskioagg_bucket,priority:4" json:"device"` + MaxReadBytesRate uint64 `json:"maxReadBytesRate"` + MaxWriteBytesRate uint64 `json:"maxWriteBytesRate"` + MaxIopsInProgress uint64 `json:"maxIopsInProgress"` +} + +func (AggregatedDiskIOMetricModel) TableName() string { + return "disk_io_metrics_aggs" +} + +// AggregatedGPUMetricModel GPU聚合表 +type AggregatedGPUMetricModel struct { + ID uint `gorm:"primaryKey;autoIncrement" json:"id"` + AgentID string `gorm:"index:idx_gpuagg_agent_bucket_idx,priority:1;uniqueIndex:ux_gpuagg_bucket,priority:1" json:"agentId"` + BucketSeconds int `gorm:"index:idx_gpuagg_agent_bucket_idx,priority:2;uniqueIndex:ux_gpuagg_bucket,priority:2" json:"bucketSeconds"` + BucketStart int64 `gorm:"index:idx_gpuagg_agent_bucket_idx,priority:3;uniqueIndex:ux_gpuagg_bucket,priority:3" json:"bucketStart"` // 毫秒 + Index int `gorm:"index:idx_gpuagg_agent_bucket_idx,priority:4;uniqueIndex:ux_gpuagg_bucket,priority:4" json:"index"` + Name string `json:"name"` + MaxUtilization float64 `json:"maxUtilization"` + MaxMemoryUsed uint64 `json:"maxMemoryUsed"` + MaxTemperature float64 `json:"maxTemperature"` + MaxPowerDraw float64 `json:"maxPowerDraw"` + MemoryTotal uint64 `json:"memoryTotal"` +} + +func (AggregatedGPUMetricModel) TableName() string { + return "gpu_metrics_aggs" +} + +// AggregatedTemperatureMetricModel 温度聚合表 +type AggregatedTemperatureMetricModel struct { + ID uint `gorm:"primaryKey;autoIncrement" json:"id"` + AgentID string `gorm:"index:idx_tempagg_agent_bucket_sensor,priority:1;uniqueIndex:ux_tempagg_bucket,priority:1" json:"agentId"` + BucketSeconds int `gorm:"index:idx_tempagg_agent_bucket_sensor,priority:2;uniqueIndex:ux_tempagg_bucket,priority:2" json:"bucketSeconds"` + BucketStart int64 `gorm:"index:idx_tempagg_agent_bucket_sensor,priority:3;uniqueIndex:ux_tempagg_bucket,priority:3" json:"bucketStart"` // 毫秒 + SensorKey string `gorm:"index:idx_tempagg_agent_bucket_sensor,priority:4;uniqueIndex:ux_tempagg_bucket,priority:4" json:"sensorKey"` + SensorLabel string `json:"sensorLabel"` + MaxTemperature float64 `json:"maxTemperature"` +} + +func (AggregatedTemperatureMetricModel) TableName() string { + return "temperature_metrics_aggs" +} + +// AggregationProgress 聚合进度记录 +type AggregationProgress struct { + MetricType string `gorm:"primaryKey"` // cpu/memory/disk/network + BucketSeconds int `gorm:"primaryKey"` + LastBucket int64 `json:"lastBucket"` // 已完成的最后 bucket 起始时间（毫秒） + UpdatedAt int64 `json:"updatedAt" gorm:"autoUpdateTime:milli"` + CreatedAt int64 `json:"createdAt" gorm:"autoCreateTime:milli"` +} + +func (AggregationProgress) TableName() string { + return "aggregation_progress" +} diff --git a/internal/models/property.go b/internal/models/property.go index ba9b144..96370e0 100644 --- a/internal/models/property.go +++ b/internal/models/property.go @@ -48,3 +48,9 @@ type SystemConfig struct { ICPCode string `json:"icpCode"` // ICP备案号 DefaultView string `json:"defaultView"` // 默认视图 grid | list } + +// MetricsConfig 指标数据配置 +type MetricsConfig struct { + RetentionHours int `json:"retentionHours"` // 原始数据保留小时数（默认168小时=7天） + MaxQueryPoints int `json:"maxQueryPoints"` // 查询返回的最大数据点数（默认720） +} diff --git a/internal/protocol/message.go b/internal/protocol/message.go index 906ba8e..cb3fe0e 100644 --- a/internal/protocol/message.go +++ b/internal/protocol/message.go @@ -66,7 +66,6 @@ const ( MetricTypeDiskIO MetricType = "disk_io" MetricTypeNetwork MetricType = "network" MetricTypeNetworkConnection MetricType = "network_connection" - MetricTypeLoad MetricType = "load" MetricTypeHost MetricType = "host" MetricTypeGPU MetricType = "gpu" MetricTypeTemperature MetricType = "temperature" diff --git a/internal/repo/metric_repo.go b/internal/repo/metric_repo.go index 8b98f2f..4fe3cdb 100644 --- a/internal/repo/metric_repo.go +++ b/internal/repo/metric_repo.go @@ -39,11 +39,6 @@ func (r *MetricRepo) SaveNetworkMetric(ctx context.Context, metric *models.Netwo return r.db.WithContext(ctx).Create(metric).Error } -// SaveLoadMetric 保存负载指标 -func (r *MetricRepo) SaveLoadMetric(ctx context.Context, metric *models.LoadMetric) error { - return r.db.WithContext(ctx).Create(metric).Error -} - // SaveDiskIOMetric 保存磁盘IO指标 func (r *MetricRepo) SaveDiskIOMetric(ctx context.Context, metric *models.DiskIOMetric) error { return r.db.WithContext(ctx).Create(metric).Error @@ -202,7 +197,6 @@ func (r *MetricRepo) DeleteOldMetrics(ctx context.Context, beforeTimestamp int64 &models.DiskMetric{}, &models.NetworkMetric{}, &models.NetworkConnectionMetric{}, - &models.LoadMetric{}, &models.DiskIOMetric{}, &models.GPUMetric{}, &models.TemperatureMetric{}, @@ -325,9 +319,10 @@ func (r *MetricRepo) GetDiskMetrics(ctx context.Context, agentID string, start, return metrics, err } -// AggregatedNetworkMetric 网络聚合指标（使用最大值） +// AggregatedNetworkMetric 网络聚合指标（按网卡分组） type AggregatedNetworkMetric struct { Timestamp int64 `json:"timestamp"` + Interface string `json:"interface"` MaxSentRate float64 `json:"maxSentRate"` MaxRecvRate float64 `json:"maxRecvRate"` } @@ -387,38 +382,6 @@ func (r *MetricRepo) GetNetworkMetricsByInterface(ctx context.Context, agentID s return metrics, err } -// AggregatedLoadMetric 负载聚合指标（使用最大值） -type AggregatedLoadMetric struct { - Timestamp int64 `json:"timestamp"` - MaxLoad1 float64 `json:"maxLoad1"` - MaxLoad5 float64 `json:"maxLoad5"` - MaxLoad15 float64 `json:"maxLoad15"` -} - -// GetLoadMetrics 获取聚合后的负载指标（始终返回聚合数据） -func (r *MetricRepo) GetLoadMetrics(ctx context.Context, agentID string, start, end int64, interval int) ([]AggregatedLoadMetric, error) { - var metrics []AggregatedLoadMetric - - query := ` - SELECT - CAST(FLOOR(timestamp / ?) * ? AS BIGINT) as timestamp, - MAX(load1) as max_load1, - MAX(load5) as max_load5, - MAX(load15) as max_load15 - FROM load_metrics - WHERE agent_id = ? AND timestamp >= ? AND timestamp <= ? - GROUP BY 1 - ORDER BY timestamp ASC - ` - - intervalMs := int64(interval * 1000) - err := r.db.WithContext(ctx). - Raw(query, intervalMs, intervalMs, agentID, start, end). - Scan(&metrics).Error - - return metrics, err -} - // GetLatestCPUMetric 获取最新的CPU指标 func (r *MetricRepo) GetLatestCPUMetric(ctx context.Context, agentID string) (*models.CPUMetric, error) { var metric models.CPUMetric @@ -499,19 +462,6 @@ func (r *MetricRepo) GetLatestNetworkMetrics(ctx context.Context, agentID string return metrics, err } -// GetLatestLoadMetric 获取最新的负载指标 -func (r *MetricRepo) GetLatestLoadMetric(ctx context.Context, agentID string) (*models.LoadMetric, error) { - var metric models.LoadMetric - err := r.db.WithContext(ctx). - Where("agent_id = ?", agentID). - Order("timestamp DESC"). - First(&metric).Error - if err != nil { - return nil, err - } - return &metric, nil -} - // SaveMonitorMetric 保存监控指标 func (r *MetricRepo) SaveMonitorMetric(ctx context.Context, metric *models.MonitorMetric) error { return r.db.WithContext(ctx).Create(metric).Error @@ -622,12 +572,13 @@ func (r *MetricRepo) GetAggregatedMonitorMetrics(ctx context.Context, monitorID // AggregatedDiskIOMetric 磁盘IO聚合指标 type AggregatedDiskIOMetric struct { - Timestamp int64 `json:"timestamp"` - Device string `json:"device"` - MaxReadRate float64 `json:"maxReadRate"` // 最大读取速率(字节/秒) - MaxWriteRate float64 `json:"maxWriteRate"` // 最大写入速率(字节/秒) - TotalReadBytes uint64 `json:"totalReadBytes"` // 总读取字节数 - TotalWriteBytes uint64 `json:"totalWriteBytes"` // 总写入字节数 + Timestamp int64 `json:"timestamp"` + Device string `json:"device"` + MaxReadRate float64 `json:"maxReadRate"` // 最大读取速率(字节/秒) + MaxWriteRate float64 `json:"maxWriteRate"` // 最大写入速率(字节/秒) + TotalReadBytes uint64 `json:"totalReadBytes"` // 总读取字节数 + TotalWriteBytes uint64 `json:"totalWriteBytes"` // 总写入字节数 + MaxIopsInProgress uint64 `json:"maxIopsInProgress"` // 最大进行中的IO操作数 } // GetDiskIOMetrics 获取聚合后的磁盘IO指标 @@ -666,10 +617,13 @@ func (r *MetricRepo) GetDiskIOMetrics(ctx context.Context, agentID string, start // AggregatedGPUMetric GPU聚合指标（使用最大值） type AggregatedGPUMetric struct { Timestamp int64 `json:"timestamp"` + Index int `json:"index"` + Name string `json:"name"` MaxUtilization float64 `json:"maxUtilization"` MaxMemoryUsed uint64 `json:"maxMemoryUsed"` MaxTemperature float64 `json:"maxTemperature"` MaxPowerDraw float64 `json:"maxPowerDraw"` + MemoryTotal uint64 `json:"memoryTotal"` } // GetGPUMetrics 获取聚合后的GPU指标 @@ -744,7 +698,6 @@ func (r *MetricRepo) DeleteAgentMetrics(ctx context.Context, agentID string) err &models.DiskMetric{}, &models.DiskIOMetric{}, &models.NetworkMetric{}, - &models.LoadMetric{}, &models.HostMetric{}, &models.GPUMetric{}, &models.TemperatureMetric{}, @@ -798,3 +751,348 @@ func (r *MetricRepo) GetAllLatestMonitorMetrics(ctx context.Context) ([]*models. return metrics, err } + +// ----------- 聚合表操作 ----------- + +// AggregateCPUToAgg 将原始CPU数据聚合到聚合表 +func (r *MetricRepo) AggregateCPUToAgg(ctx context.Context, bucketSeconds int, start, end int64) error { + bucketMs := int64(bucketSeconds * 1000) + return r.db.WithContext(ctx).Exec(` + INSERT INTO cpu_metrics_aggs (agent_id, bucket_seconds, bucket_start, max_usage, logical_cores) + SELECT + agent_id, + ? as bucket_seconds, + (timestamp / ?) * ? as bucket_start, + MAX(usage_percent) as max_usage, + MAX(logical_cores) as logical_cores + FROM cpu_metrics + WHERE timestamp >= ? AND timestamp < ? + GROUP BY agent_id, bucket_start + ON CONFLICT (agent_id, bucket_seconds, bucket_start) DO UPDATE SET + max_usage = EXCLUDED.max_usage, + logical_cores = EXCLUDED.logical_cores + `, bucketSeconds, bucketMs, bucketMs, start, end).Error +} + +// AggregateMemoryToAgg 将原始内存数据聚合到聚合表 +func (r *MetricRepo) AggregateMemoryToAgg(ctx context.Context, bucketSeconds int, start, end int64) error { + bucketMs := int64(bucketSeconds * 1000) + return r.db.WithContext(ctx).Exec(` + INSERT INTO memory_metrics_aggs (agent_id, bucket_seconds, bucket_start, max_usage, total) + SELECT + agent_id, + ? as bucket_seconds, + (timestamp / ?) * ? as bucket_start, + MAX(usage_percent) as max_usage, + MAX(total) as total + FROM memory_metrics + WHERE timestamp >= ? AND timestamp < ? + GROUP BY agent_id, bucket_start + ON CONFLICT (agent_id, bucket_seconds, bucket_start) DO UPDATE SET + max_usage = EXCLUDED.max_usage, + total = EXCLUDED.total + `, bucketSeconds, bucketMs, bucketMs, start, end).Error +} + +// AggregateDiskToAgg 将原始磁盘数据聚合到聚合表 +func (r *MetricRepo) AggregateDiskToAgg(ctx context.Context, bucketSeconds int, start, end int64) error { + bucketMs := int64(bucketSeconds * 1000) + return r.db.WithContext(ctx).Exec(` + INSERT INTO disk_metrics_aggs (agent_id, bucket_seconds, bucket_start, mount_point, max_usage, total) + SELECT + agent_id, + ? as bucket_seconds, + (timestamp / ?) * ? as bucket_start, + mount_point, + MAX(usage_percent) as max_usage, + MAX(total) as total + FROM disk_metrics + WHERE timestamp >= ? AND timestamp < ? + GROUP BY agent_id, bucket_start, mount_point + ON CONFLICT (agent_id, bucket_seconds, bucket_start, mount_point) DO UPDATE SET + max_usage = EXCLUDED.max_usage, + total = EXCLUDED.total + `, bucketSeconds, bucketMs, bucketMs, start, end).Error +} + +// AggregateNetworkToAgg 将原始网络数据聚合到聚合表（按网卡分组） +func (r *MetricRepo) AggregateNetworkToAgg(ctx context.Context, bucketSeconds int, start, end int64) error { + bucketMs := int64(bucketSeconds * 1000) + return r.db.WithContext(ctx).Exec(` + INSERT INTO network_metrics_aggs (agent_id, bucket_seconds, bucket_start, interface, max_sent_rate, max_recv_rate) + SELECT + agent_id, + ? as bucket_seconds, + (timestamp / ?) * ? as bucket_start, + interface, + MAX(bytes_sent_rate) as max_sent_rate, + MAX(bytes_recv_rate) as max_recv_rate + FROM network_metrics + WHERE timestamp >= ? AND timestamp < ? + GROUP BY agent_id, bucket_start, interface + ON CONFLICT (agent_id, bucket_seconds, bucket_start, interface) DO UPDATE SET + max_sent_rate = EXCLUDED.max_sent_rate, + max_recv_rate = EXCLUDED.max_recv_rate + `, bucketSeconds, bucketMs, bucketMs, start, end).Error +} + +// AggregateNetworkConnectionToAgg 将原始网络连接数据聚合到聚合表 +func (r *MetricRepo) AggregateNetworkConnectionToAgg(ctx context.Context, bucketSeconds int, start, end int64) error { + bucketMs := int64(bucketSeconds * 1000) + return r.db.WithContext(ctx).Exec(` + INSERT INTO network_connection_metrics_aggs ( + agent_id, bucket_seconds, bucket_start, + max_established, max_syn_sent, max_syn_recv, max_fin_wait1, max_fin_wait2, + max_time_wait, max_close, max_close_wait, max_last_ack, max_listen, + max_closing, max_total + ) + SELECT + agent_id, + ? as bucket_seconds, + (timestamp / ?) * ? as bucket_start, + MAX(established) as max_established, + MAX(syn_sent) as max_syn_sent, + MAX(syn_recv) as max_syn_recv, + MAX(fin_wait1) as max_fin_wait1, + MAX(fin_wait2) as max_fin_wait2, + MAX(time_wait) as max_time_wait, + MAX(close) as max_close, + MAX(close_wait) as max_close_wait, + MAX(last_ack) as max_last_ack, + MAX(listen) as max_listen, + MAX(closing) as max_closing, + MAX(total) as max_total + FROM network_connection_metrics + WHERE timestamp >= ? AND timestamp < ? + GROUP BY agent_id, bucket_start + ON CONFLICT (agent_id, bucket_seconds, bucket_start) DO UPDATE SET + max_established = EXCLUDED.max_established, + max_syn_sent = EXCLUDED.max_syn_sent, + max_syn_recv = EXCLUDED.max_syn_recv, + max_fin_wait1 = EXCLUDED.max_fin_wait1, + max_fin_wait2 = EXCLUDED.max_fin_wait2, + max_time_wait = EXCLUDED.max_time_wait, + max_close = EXCLUDED.max_close, + max_close_wait = EXCLUDED.max_close_wait, + max_last_ack = EXCLUDED.max_last_ack, + max_listen = EXCLUDED.max_listen, + max_closing = EXCLUDED.max_closing, + max_total = EXCLUDED.max_total + `, bucketSeconds, bucketMs, bucketMs, start, end).Error +} + +// AggregateDiskIOToAgg 将原始磁盘IO数据聚合到聚合表 +func (r *MetricRepo) AggregateDiskIOToAgg(ctx context.Context, bucketSeconds int, start, end int64) error { + bucketMs := int64(bucketSeconds * 1000) + return r.db.WithContext(ctx).Exec(` + INSERT INTO disk_io_metrics_aggs ( + agent_id, bucket_seconds, bucket_start, device, + max_read_bytes_rate, max_write_bytes_rate, max_iops_in_progress + ) + SELECT + agent_id, + ? as bucket_seconds, + (timestamp / ?) * ? as bucket_start, + device, + MAX(read_bytes_rate) as max_read_bytes_rate, + MAX(write_bytes_rate) as max_write_bytes_rate, + MAX(iops_in_progress) as max_iops_in_progress + FROM disk_io_metrics + WHERE timestamp >= ? AND timestamp < ? + GROUP BY agent_id, bucket_start, device + ON CONFLICT (agent_id, bucket_seconds, bucket_start, device) DO UPDATE SET + max_read_bytes_rate = EXCLUDED.max_read_bytes_rate, + max_write_bytes_rate = EXCLUDED.max_write_bytes_rate, + max_iops_in_progress = EXCLUDED.max_iops_in_progress + `, bucketSeconds, bucketMs, bucketMs, start, end).Error +} + +// AggregateGPUToAgg 将原始GPU数据聚合到聚合表 +func (r *MetricRepo) AggregateGPUToAgg(ctx context.Context, bucketSeconds int, start, end int64) error { + bucketMs := int64(bucketSeconds * 1000) + return r.db.WithContext(ctx).Exec(` + INSERT INTO gpu_metrics_aggs ( + agent_id, bucket_seconds, bucket_start, index, name, + max_utilization, max_memory_used, max_temperature, max_power_draw, memory_total + ) + SELECT + agent_id, + ? as bucket_seconds, + (timestamp / ?) * ? as bucket_start, + index, + MAX(name) as name, + MAX(utilization) as max_utilization, + MAX(memory_used) as max_memory_used, + MAX(temperature) as max_temperature, + MAX(power_draw) as max_power_draw, + MAX(memory_total) as memory_total + FROM gpu_metrics + WHERE timestamp >= ? AND timestamp < ? + GROUP BY agent_id, bucket_start, index + ON CONFLICT (agent_id, bucket_seconds, bucket_start, index) DO UPDATE SET + name = EXCLUDED.name, + max_utilization = EXCLUDED.max_utilization, + max_memory_used = EXCLUDED.max_memory_used, + max_temperature = EXCLUDED.max_temperature, + max_power_draw = EXCLUDED.max_power_draw, + memory_total = EXCLUDED.memory_total + `, bucketSeconds, bucketMs, bucketMs, start, end).Error +} + +// AggregateTemperatureToAgg 将原始温度数据聚合到聚合表 +func (r *MetricRepo) AggregateTemperatureToAgg(ctx context.Context, bucketSeconds int, start, end int64) error { + bucketMs := int64(bucketSeconds * 1000) + return r.db.WithContext(ctx).Exec(` + INSERT INTO temperature_metrics_aggs ( + agent_id, bucket_seconds, bucket_start, sensor_key, sensor_label, max_temperature + ) + SELECT + agent_id, + ? as bucket_seconds, + (timestamp / ?) * ? as bucket_start, + sensor_key, + MAX(sensor_label) as sensor_label, + MAX(temperature) as max_temperature + FROM temperature_metrics + WHERE timestamp >= ? AND timestamp < ? + GROUP BY agent_id, bucket_start, sensor_key + ON CONFLICT (agent_id, bucket_seconds, bucket_start, sensor_key) DO UPDATE SET + sensor_label = EXCLUDED.sensor_label, + max_temperature = EXCLUDED.max_temperature + `, bucketSeconds, bucketMs, bucketMs, start, end).Error +} + +// GetCPUMetricsAgg 从聚合表获取CPU指标 +func (r *MetricRepo) GetCPUMetricsAgg(ctx context.Context, agentID string, start, end int64, bucketSeconds int) ([]AggregatedCPUMetric, error) { + var metrics []AggregatedCPUMetric + err := r.db.WithContext(ctx). + Table("cpu_metrics_aggs"). + Select("bucket_start as timestamp, max_usage, logical_cores"). + Where("agent_id = ? AND bucket_seconds = ? AND bucket_start >= ? AND bucket_start < ?", agentID, bucketSeconds, start, end). + Order("bucket_start"). + Scan(&metrics).Error + return metrics, err +} + +// GetMemoryMetricsAgg 从聚合表获取内存指标 +func (r *MetricRepo) GetMemoryMetricsAgg(ctx context.Context, agentID string, start, end int64, bucketSeconds int) ([]AggregatedMemoryMetric, error) { + var metrics []AggregatedMemoryMetric + err := r.db.WithContext(ctx). + Table("memory_metrics_aggs"). + Select("bucket_start as timestamp, max_usage, total"). + Where("agent_id = ? AND bucket_seconds = ? AND bucket_start >= ? AND bucket_start < ?", agentID, bucketSeconds, start, end). + Order("bucket_start"). + Scan(&metrics).Error + return metrics, err +} + +// GetDiskMetricsAgg 从聚合表获取磁盘指标 +func (r *MetricRepo) GetDiskMetricsAgg(ctx context.Context, agentID string, start, end int64, bucketSeconds int) ([]AggregatedDiskMetric, error) { + var metrics []AggregatedDiskMetric + err := r.db.WithContext(ctx). + Table("disk_metrics_aggs"). + Select("bucket_start as timestamp, mount_point, max_usage, total"). + Where("agent_id = ? AND bucket_seconds = ? AND bucket_start >= ? AND bucket_start < ?", agentID, bucketSeconds, start, end). + Order("bucket_start, mount_point"). + Scan(&metrics).Error + return metrics, err +} + +// GetNetworkMetricsAgg 从聚合表获取网络指标 +func (r *MetricRepo) GetNetworkMetricsAgg(ctx context.Context, agentID string, start, end int64, bucketSeconds int) ([]AggregatedNetworkMetric, error) { + var metrics []AggregatedNetworkMetric + err := r.db.WithContext(ctx). + Table("network_metrics_aggs"). + Select("bucket_start as timestamp, interface, max_sent_rate, max_recv_rate"). + Where("agent_id = ? AND bucket_seconds = ? AND bucket_start >= ? AND bucket_start < ?", agentID, bucketSeconds, start, end). + Order("bucket_start, interface"). + Scan(&metrics).Error + return metrics, err +} + +// GetNetworkConnectionMetricsAgg 从聚合表获取网络连接指标 +func (r *MetricRepo) GetNetworkConnectionMetricsAgg(ctx context.Context, agentID string, start, end int64, bucketSeconds int) ([]AggregatedNetworkConnectionMetric, error) { + var metrics []AggregatedNetworkConnectionMetric + err := r.db.WithContext(ctx). + Table("network_connection_metrics_aggs"). + Select(`bucket_start as timestamp, + max_established, max_syn_sent, max_syn_recv, + max_fin_wait1, max_fin_wait2, max_time_wait, + max_close, max_close_wait, max_last_ack, + max_listen, max_closing, max_total`). + Where("agent_id = ? AND bucket_seconds = ? AND bucket_start >= ? AND bucket_start < ?", agentID, bucketSeconds, start, end). + Order("bucket_start"). + Scan(&metrics).Error + return metrics, err +} + +// GetDiskIOMetricsAgg 从聚合表获取磁盘IO指标 +func (r *MetricRepo) GetDiskIOMetricsAgg(ctx context.Context, agentID string, start, end int64, bucketSeconds int) ([]AggregatedDiskIOMetric, error) { + var metrics []AggregatedDiskIOMetric + err := r.db.WithContext(ctx). + Table("disk_io_metrics_aggs"). + Select(`bucket_start as timestamp, device, + max_read_bytes_rate as max_read_rate, + max_write_bytes_rate as max_write_rate, + max_iops_in_progress`). + Where("agent_id = ? AND bucket_seconds = ? AND bucket_start >= ? AND bucket_start < ?", agentID, bucketSeconds, start, end). + Order("bucket_start, device"). + Scan(&metrics).Error + return metrics, err +} + +// GetGPUMetricsAgg 从聚合表获取GPU指标 +func (r *MetricRepo) GetGPUMetricsAgg(ctx context.Context, agentID string, start, end int64, bucketSeconds int) ([]AggregatedGPUMetric, error) { + var metrics []AggregatedGPUMetric + err := r.db.WithContext(ctx). + Table("gpu_metrics_aggs"). + Select(`bucket_start as timestamp, index, name, + max_utilization, + max_memory_used, + max_temperature, + max_power_draw, + memory_total`). + Where("agent_id = ? AND bucket_seconds = ? AND bucket_start >= ? AND bucket_start < ?", agentID, bucketSeconds, start, end). + Order("bucket_start, index"). + Scan(&metrics).Error + return metrics, err +} + +// GetTemperatureMetricsAgg 从聚合表获取温度指标 +func (r *MetricRepo) GetTemperatureMetricsAgg(ctx context.Context, agentID string, start, end int64, bucketSeconds int) ([]AggregatedTemperatureMetric, error) { + var metrics []AggregatedTemperatureMetric + err := r.db.WithContext(ctx). + Table("temperature_metrics_aggs"). + Select(`bucket_start as timestamp, sensor_key, sensor_label, + max_temperature`). + Where("agent_id = ? AND bucket_seconds = ? AND bucket_start >= ? AND bucket_start < ?", agentID, bucketSeconds, start, end). + Order("bucket_start, sensor_key"). + Scan(&metrics).Error + return metrics, err +} + +// UpsertAggregationProgress 更新或插入聚合进度 +func (r *MetricRepo) UpsertAggregationProgress(ctx context.Context, metricType string, bucketSeconds int, lastBucket int64) error { + return r.db.WithContext(ctx). + Clauses(clause.OnConflict{ + Columns: []clause.Column{{Name: "metric_type"}, {Name: "bucket_seconds"}}, + DoUpdates: clause.AssignmentColumns([]string{"last_bucket"}), + }). + Create(&models.AggregationProgress{ + MetricType: metricType, + BucketSeconds: bucketSeconds, + LastBucket: lastBucket, + }).Error +} + +// GetAggregationProgress 获取聚合进度 +func (r *MetricRepo) GetAggregationProgress(ctx context.Context, metricType string, bucketSeconds int) (*models.AggregationProgress, error) { + var progress models.AggregationProgress + err := r.db.WithContext(ctx). + Where("metric_type = ? AND bucket_seconds = ?", metricType, bucketSeconds). + First(&progress).Error + if err != nil { + return nil, err + } + return &progress, nil +} diff --git a/internal/service/agent_service.go b/internal/service/agent_service.go index dd104ec..f7fe5a4 100644 --- a/internal/service/agent_service.go +++ b/internal/service/agent_service.go @@ -5,6 +5,7 @@ import ( "encoding/json" "errors" "fmt" + "math" "time" "github.com/dushixiang/pika/internal/models" @@ -15,6 +16,19 @@ import ( "gorm.io/gorm" ) +const ( + defaultMetricsRetentionHours = 24 * 7 // 默认保留 7 天 + defaultMaxQueryPoints = 720 // 默认最多返回 720 个点 +) + +var allowedIntervals = []int{ + 1, 2, 5, + 10, 15, 30, + 60, 120, 300, + 600, 900, 1800, + 3600, 7200, 14400, +} + type AgentService struct { logger *zap.Logger *orz.Service @@ -22,9 +36,10 @@ type AgentService struct { metricRepo *repo.MetricRepo monitorStatsRepo *repo.MonitorStatsRepo apiKeyService *ApiKeyService + propertyService *PropertyService } -func NewAgentService(logger *zap.Logger, db *gorm.DB, apiKeyService *ApiKeyService) *AgentService { +func NewAgentService(logger *zap.Logger, db *gorm.DB, apiKeyService *ApiKeyService, propertyService *PropertyService) *AgentService { return &AgentService{ logger: logger, Service: orz.NewService(db), @@ -32,6 +47,7 @@ func NewAgentService(logger *zap.Logger, db *gorm.DB, apiKeyService *ApiKeyServi metricRepo: repo.NewMetricRepo(db), monitorStatsRepo: repo.NewMonitorStatsRepo(db), apiKeyService: apiKeyService, + propertyService: propertyService, } } @@ -246,20 +262,6 @@ func (s *AgentService) HandleMetricData(ctx context.Context, agentID string, met } return s.metricRepo.SaveNetworkConnectionMetric(ctx, metric) - case protocol.MetricTypeLoad: - var loadData protocol.LoadData - if err := json.Unmarshal(data, &loadData); err != nil { - return err - } - metric := &models.LoadMetric{ - AgentID: agentID, - Load1: loadData.Load1, - Load5: loadData.Load5, - Load15: loadData.Load15, - Timestamp: now, - } - return s.metricRepo.SaveLoadMetric(ctx, metric) - case protocol.MetricTypeDiskIO: // DiskIO现在是数组,需要批量处理 var diskIODataList []*protocol.DiskIOData @@ -424,26 +426,96 @@ func CalculateInterval(start, end int64) int { } } -// GetMetrics 获取聚合指标数据 +// GetMetrics 获取聚合指标数据（自动路由到聚合表或原始表） func (s *AgentService) GetMetrics(ctx context.Context, agentID, metricType string, start, end int64, interval int) (interface{}, error) { + start, end = s.normalizeTimeRange(ctx, start, end) + interval = s.DetermineInterval(ctx, start, end, interval) + + // 判断是否可以使用聚合表（仅支持部分指标类型） + aggCapable := map[string]bool{ + "cpu": true, + "memory": true, + "disk": true, + "network": true, + "network_connection": true, + "disk_io": true, + "gpu": true, + "temperature": true, + } + + // 智能选择聚合粒度：根据查询间隔选择最合适的bucket + // 例如：查询90秒数据时使用60秒bucket，查询600秒数据时使用300秒bucket + var bucketSeconds int + useAgg := false + if aggCapable[metricType] { + if interval >= 3600 { + bucketSeconds = 3600 + useAgg = true + } else if interval >= 300 { + bucketSeconds = 300 + useAgg = true + } else if interval >= 60 { + bucketSeconds = 60 + useAgg = true + } + } + switch metricType { case "cpu": + if useAgg { + if metrics, err := s.metricRepo.GetCPUMetricsAgg(ctx, agentID, start, end, bucketSeconds); err == nil && len(metrics) > 0 { + return metrics, nil + } + } return s.metricRepo.GetCPUMetrics(ctx, agentID, start, end, interval) case "memory": + if useAgg { + if metrics, err := s.metricRepo.GetMemoryMetricsAgg(ctx, agentID, start, end, bucketSeconds); err == nil && len(metrics) > 0 { + return metrics, nil + } + } return s.metricRepo.GetMemoryMetrics(ctx, agentID, start, end, interval) case "disk": + if useAgg { + if metrics, err := s.metricRepo.GetDiskMetricsAgg(ctx, agentID, start, end, bucketSeconds); err == nil && len(metrics) > 0 { + return metrics, nil + } + } return s.metricRepo.GetDiskMetrics(ctx, agentID, start, end, interval) case "network": + if useAgg { + if metrics, err := s.metricRepo.GetNetworkMetricsAgg(ctx, agentID, start, end, bucketSeconds); err == nil && len(metrics) > 0 { + return metrics, nil + } + } return s.metricRepo.GetNetworkMetrics(ctx, agentID, start, end, interval) case "network_connection": + if useAgg { + if metrics, err := s.metricRepo.GetNetworkConnectionMetricsAgg(ctx, agentID, start, end, bucketSeconds); err == nil && len(metrics) > 0 { + return metrics, nil + } + } return s.metricRepo.GetNetworkConnectionMetrics(ctx, agentID, start, end, interval) - case "load": - return s.metricRepo.GetLoadMetrics(ctx, agentID, start, end, interval) case "disk_io": + if useAgg { + if metrics, err := s.metricRepo.GetDiskIOMetricsAgg(ctx, agentID, start, end, bucketSeconds); err == nil && len(metrics) > 0 { + return metrics, nil + } + } return s.metricRepo.GetDiskIOMetrics(ctx, agentID, start, end, interval) case "gpu": + if useAgg { + if metrics, err := s.metricRepo.GetGPUMetricsAgg(ctx, agentID, start, end, bucketSeconds); err == nil && len(metrics) > 0 { + return metrics, nil + } + } return s.metricRepo.GetGPUMetrics(ctx, agentID, start, end, interval) case "temperature": + if useAgg { + if metrics, err := s.metricRepo.GetTemperatureMetricsAgg(ctx, agentID, start, end, bucketSeconds); err == nil && len(metrics) > 0 { + return metrics, nil + } + } return s.metricRepo.GetTemperatureMetrics(ctx, agentID, start, end, interval) default: return nil, nil @@ -452,9 +524,187 @@ func (s *AgentService) GetMetrics(ctx context.Context, agentID, metricType strin // GetNetworkMetricsByInterface 获取按网卡接口分组的网络指标 func (s *AgentService) GetNetworkMetricsByInterface(ctx context.Context, agentID string, start, end int64, interval int) (interface{}, error) { + start, end = s.normalizeTimeRange(ctx, start, end) + interval = s.DetermineInterval(ctx, start, end, interval) return s.metricRepo.GetNetworkMetricsByInterface(ctx, agentID, start, end, interval) } +// DetermineInterval 根据配置、用户请求和时间范围决定聚合粒度 +func (s *AgentService) DetermineInterval(ctx context.Context, start, end int64, requested int) int { + cfg := s.getMetricsConfig(ctx) + + interval := requested + if interval <= 0 { + interval = calculateBaseInterval(start, end) + } + + interval = adjustIntervalForMaxPoints(start, end, interval, cfg.MaxQueryPoints) + return interval +} + +// normalizeTimeRange 将时间范围限制在保留周期内，避免无意义的全表扫描 +func (s *AgentService) normalizeTimeRange(ctx context.Context, start, end int64) (int64, int64) { + cfg := s.getMetricsConfig(ctx) + retentionDuration := time.Duration(cfg.RetentionHours) * time.Hour + + retentionBoundary := time.Now().Add(-retentionDuration).UnixMilli() + if start < retentionBoundary { + start = retentionBoundary + } + if end <= start { + end = start + 1000 + } + return start, end +} + +// 默认间隔选择（保留旧行为的同时限制最大数据点） +func defaultIntervalWithLimit(start, end int64) int { + base := calculateBaseInterval(start, end) + return adjustIntervalForMaxPoints(start, end, base, defaultMaxQueryPoints) +} + +func calculateBaseInterval(start, end int64) int { + duration := (end - start) / 1000 // 秒 + + switch { + case duration <= 60: // 1 分钟内 + return 2 + case duration <= 5*60: // 5 分钟内 + return 5 + case duration <= 15*60: // 15 分钟内 + return 15 + case duration <= 30*60: // 30 分钟内 + return 30 + case duration <= 60*60: // 1 小时内 + return 60 + case duration <= 3*60*60: // 3 小时内 + return 180 + case duration <= 6*60*60: // 6 小时内 + return 300 + case duration <= 12*60*60: // 12 小时内 + return 600 + case duration <= 24*60*60: // 1 天内 + return 900 + case duration <= 3*24*60*60: // 3 天内 + return 1800 + case duration <= 7*24*60*60: // 7 天内 + return 3600 + case duration <= 14*24*60*60: // 14 天内 + return 7200 + default: + return 14400 // 更长时间：4 小时粒度 + } +} + +// adjustIntervalForMaxPoints 根据最大数据点限制提升聚合粒度 +func adjustIntervalForMaxPoints(start, end int64, interval int, maxPoints int) int { + if interval <= 0 { + interval = 1 + } + if maxPoints <= 0 { + return alignInterval(interval) + } + + durationSeconds := float64(end-start) / 1000 + if durationSeconds <= 0 { + return alignInterval(interval) + } + + required := int(math.Ceil(durationSeconds / float64(maxPoints))) + interval = maxInt(interval, required) + return alignInterval(interval) +} + +func alignInterval(interval int) int { + for _, candidate := range allowedIntervals { + if interval <= candidate { + return candidate + } + } + return allowedIntervals[len(allowedIntervals)-1] +} + +func maxInt(a, b int) int { + if a > b { + return a + } + return b +} + +func (s *AgentService) getMetricsConfig(ctx context.Context) models.MetricsConfig { + cfg := models.MetricsConfig{ + RetentionHours: defaultMetricsRetentionHours, + MaxQueryPoints: defaultMaxQueryPoints, + } + + if s.propertyService == nil { + return cfg + } + + loaded := s.propertyService.GetMetricsConfig(ctx) + if loaded.RetentionHours > 0 { + cfg.RetentionHours = loaded.RetentionHours + } + if loaded.MaxQueryPoints > 0 { + cfg.MaxQueryPoints = loaded.MaxQueryPoints + } + return cfg +} + +// runAggregation 按固定 bucket 下采样存储 +func (s *AgentService) runAggregation(ctx context.Context) { + cfg := s.getMetricsConfig(ctx) + retention := time.Duration(cfg.RetentionHours) * time.Hour + // 标准 bucket：1m、5m、1h + buckets := []int{60, 300, 3600} + + for _, bucket := range buckets { + s.aggregateMetric(ctx, "cpu", bucket, retention, s.metricRepo.AggregateCPUToAgg) + s.aggregateMetric(ctx, "memory", bucket, retention, s.metricRepo.AggregateMemoryToAgg) + s.aggregateMetric(ctx, "disk", bucket, retention, s.metricRepo.AggregateDiskToAgg) + s.aggregateMetric(ctx, "network", bucket, retention, s.metricRepo.AggregateNetworkToAgg) + s.aggregateMetric(ctx, "network_connection", bucket, retention, s.metricRepo.AggregateNetworkConnectionToAgg) + s.aggregateMetric(ctx, "disk_io", bucket, retention, s.metricRepo.AggregateDiskIOToAgg) + s.aggregateMetric(ctx, "gpu", bucket, retention, s.metricRepo.AggregateGPUToAgg) + s.aggregateMetric(ctx, "temperature", bucket, retention, s.metricRepo.AggregateTemperatureToAgg) + } +} + +type aggregateFn func(ctx context.Context, bucketSeconds int, start, end int64) error + +func (s *AgentService) aggregateMetric(ctx context.Context, metricType string, bucketSeconds int, retention time.Duration, fn aggregateFn) { + bucketMs := int64(bucketSeconds * 1000) + + start := s.getAggregationStart(ctx, metricType, bucketSeconds, retention, bucketMs) + endBucket := (time.Now().Add(-time.Duration(bucketSeconds)*time.Second).UnixMilli() / bucketMs) * bucketMs + + if endBucket <= start { + return + } + + end := endBucket + bucketMs - 1 + + if err := fn(ctx, bucketSeconds, start, end); err != nil { + s.logger.Error("aggregate metric failed", zap.String("metricType", metricType), zap.Int("bucketSeconds", bucketSeconds), zap.Error(err)) + return + } + + if err := s.metricRepo.UpsertAggregationProgress(ctx, metricType, bucketSeconds, endBucket); err != nil { + s.logger.Error("update aggregation progress failed", zap.String("metricType", metricType), zap.Int("bucketSeconds", bucketSeconds), zap.Error(err)) + } +} + +func (s *AgentService) getAggregationStart(ctx context.Context, metricType string, bucketSeconds int, retention time.Duration, bucketMs int64) int64 { + progress, err := s.metricRepo.GetAggregationProgress(ctx, metricType, bucketSeconds) + if err == nil && progress != nil && progress.LastBucket > 0 { + return progress.LastBucket + bucketMs + } + + // 默认从保留窗口开始 + start := time.Now().Add(-retention).UnixMilli() + return (start / bucketMs) * bucketMs +} + // GetLatestMetrics 获取最新指标 func (s *AgentService) GetLatestMetrics(ctx context.Context, agentID string) (*LatestMetrics, error) { result := &LatestMetrics{} @@ -507,11 +757,6 @@ func (s *AgentService) GetLatestMetrics(ctx context.Context, agentID string) (*L } } - // 获取最新负载信息 - if load, err := s.metricRepo.GetLatestLoadMetric(ctx, agentID); err == nil { - result.Load = load - } - // 获取最新主机信息 if host, err := s.metricRepo.GetLatestHostMetric(ctx, agentID); err == nil { result.Host = host @@ -555,10 +800,11 @@ func (s *AgentService) StartCleanupTask(ctx context.Context) { // cleanupOldMetrics 清理旧数据 func (s *AgentService) cleanupOldMetrics(ctx context.Context) { - // 删除1小时前的数据 - before := time.Now().Add(-1 * time.Hour).UnixMilli() + cfg := s.getMetricsConfig(ctx) + retentionDuration := time.Duration(cfg.RetentionHours) * time.Hour + before := time.Now().Add(-retentionDuration).UnixMilli() - s.logger.Info("starting to clean old metrics", zap.Int64("beforeTimestamp", before)) + s.logger.Info("starting to clean old metrics", zap.Int64("beforeTimestamp", before), zap.Int("retentionHours", cfg.RetentionHours)) if err := s.metricRepo.DeleteOldMetrics(ctx, before); err != nil { s.logger.Error("failed to clean old metrics", zap.Error(err)) @@ -568,6 +814,24 @@ func (s *AgentService) cleanupOldMetrics(ctx context.Context) { s.logger.Info("old metrics cleaned successfully") } +// StartAggregationTask 启动聚合下采样任务 +func (s *AgentService) StartAggregationTask(ctx context.Context) { + ticker := time.NewTicker(1 * time.Minute) + defer ticker.Stop() + + s.logger.Info("aggregation task started") + + for { + select { + case <-ctx.Done(): + s.logger.Info("aggregation task stopped") + return + case <-ticker.C: + s.runAggregation(ctx) + } + } +} + // DiskSummary 磁盘汇总数据 type DiskSummary struct { AvgUsagePercent float64 `json:"avgUsagePercent"` // 平均使用率 @@ -593,7 +857,6 @@ type LatestMetrics struct { Disk *DiskSummary `json:"disk,omitempty"` Network *NetworkSummary `json:"network,omitempty"` NetworkConnection *models.NetworkConnectionMetric `json:"networkConnection,omitempty"` - Load *models.LoadMetric `json:"load,omitempty"` Host *models.HostMetric `json:"host,omitempty"` GPU []models.GPUMetric `json:"gpu,omitempty"` Temp []models.TemperatureMetric `json:"temperature,omitempty"` diff --git a/internal/service/property_service.go b/internal/service/property_service.go index 28ed713..466f71f 100644 --- a/internal/service/property_service.go +++ b/internal/service/property_service.go @@ -19,6 +19,8 @@ const ( PropertyIDNotificationChannels = "notification_channels" // PropertyIDSystemConfig 系统配置的固定 ID PropertyIDSystemConfig = "system_config" + // PropertyIDMetricsConfig 指标配置的固定 ID + PropertyIDMetricsConfig = "metrics_config" ) type PropertyService struct { @@ -123,6 +125,22 @@ func (s *PropertyService) GetSystemConfig(ctx context.Context) (*models.SystemCo return &systemConfig, nil } +// GetMetricsConfig 获取指标配置 +func (s *PropertyService) GetMetricsConfig(ctx context.Context) models.MetricsConfig { + var config models.MetricsConfig + err := s.GetValue(ctx, PropertyIDMetricsConfig, &config) + if err != nil { + // 返回默认配置 + return models.MetricsConfig{} + } + return config +} + +// SetMetricsConfig 设置指标配置 +func (s *PropertyService) SetMetricsConfig(ctx context.Context, config models.MetricsConfig) error { + return s.Set(ctx, PropertyIDMetricsConfig, "指标数据配置", config) +} + // defaultPropertyConfig 默认配置项定义 type defaultPropertyConfig struct { ID string @@ -150,6 +168,14 @@ func (s *PropertyService) InitializeDefaultConfigs(ctx context.Context) error { Name: "通知渠道配置", Value: []models.NotificationChannelConfig{}, }, + { + ID: PropertyIDMetricsConfig, + Name: "指标数据配置", + Value: models.MetricsConfig{ + RetentionHours: 168, // 默认7天 + MaxQueryPoints: 720, // 默认720个点 + }, + }, } // 遍历并初始化每个配置 diff --git a/internal/wire_gen.go b/internal/wire_gen.go index 3c3de95..0d6cea6 100644 --- a/internal/wire_gen.go +++ b/internal/wire_gen.go @@ -26,14 +26,14 @@ func InitializeApp(logger *zap.Logger, db *gorm.DB, cfg *config.AppConfig) (*App accountService := service.NewAccountService(logger, userService, oidcService, gitHubOAuthService, cfg) accountHandler := handler.NewAccountHandler(accountService) apiKeyService := service.NewApiKeyService(logger, db) - agentService := service.NewAgentService(logger, db, apiKeyService) + propertyService := service.NewPropertyService(logger, db) + agentService := service.NewAgentService(logger, db, apiKeyService, propertyService) manager := websocket.NewManager(logger) monitorService := service.NewMonitorService(logger, db, manager) tamperRepo := repo.NewTamperRepo(db) tamperService := service.NewTamperService(logger, tamperRepo, manager) agentHandler := handler.NewAgentHandler(logger, agentService, monitorService, tamperService, manager) apiKeyHandler := handler.NewApiKeyHandler(logger, apiKeyService) - propertyService := service.NewPropertyService(logger, db) notifier := service.NewNotifier(logger) alertService := service.NewAlertService(logger, db, propertyService, notifier) alertHandler := handler.NewAlertHandler(logger, alertService) diff --git a/pkg/agent/collector/manager.go b/pkg/agent/collector/manager.go index f2b9a7a..413274f 100644 --- a/pkg/agent/collector/manager.go +++ b/pkg/agent/collector/manager.go @@ -20,7 +20,6 @@ type Manager struct { diskIOCollector *DiskIOCollector networkCollector *NetworkCollector networkConnectionCollector *NetworkConnectionCollector - loadCollector *LoadCollector hostCollector *HostCollector temperatureCollector *TemperatureCollector gpuCollector *GPUCollector @@ -36,7 +35,6 @@ func NewManager(cfg *config.Config) *Manager { diskIOCollector: NewDiskIOCollector(), networkCollector: NewNetworkCollector(cfg), networkConnectionCollector: NewNetworkConnectionCollector(), - loadCollector: NewLoadCollector(), hostCollector: NewHostCollector(), temperatureCollector: NewTemperatureCollector(), gpuCollector: NewGPUCollector(), @@ -100,15 +98,6 @@ func (m *Manager) CollectAndSendNetworkConnection(conn WebSocketWriter) error { return m.sendMetrics(conn, protocol.MetricTypeNetworkConnection, connectionData) } -// CollectAndSendLoad 采集并发送系统负载指标 -func (m *Manager) CollectAndSendLoad(conn WebSocketWriter) error { - loadData, err := m.loadCollector.Collect() - if err != nil { - return err - } - return m.sendMetrics(conn, protocol.MetricTypeLoad, loadData) -} - // CollectAndSendHost 采集并发送主机信息 func (m *Manager) CollectAndSendHost(conn WebSocketWriter) error { hostData, err := m.hostCollector.Collect() diff --git a/pkg/agent/collector/system.go b/pkg/agent/collector/system.go index cc17195..3069b4b 100644 --- a/pkg/agent/collector/system.go +++ b/pkg/agent/collector/system.go @@ -3,31 +3,8 @@ package collector import ( "github.com/dushixiang/pika/internal/protocol" "github.com/shirou/gopsutil/v4/host" - "github.com/shirou/gopsutil/v4/load" ) -// LoadCollector 系统负载监控采集器 -type LoadCollector struct{} - -// NewLoadCollector 创建负载采集器 -func NewLoadCollector() *LoadCollector { - return &LoadCollector{} -} - -// Collect 采集系统负载数据 -func (l *LoadCollector) Collect() (*protocol.LoadData, error) { - avgStat, err := load.Avg() - if err != nil { - return nil, err - } - - return &protocol.LoadData{ - Load1: avgStat.Load1, - Load5: avgStat.Load5, - Load15: avgStat.Load15, - }, nil -} - // HostCollector 主机信息采集器 type HostCollector struct{} diff --git a/pkg/agent/service/agent.go b/pkg/agent/service/agent.go index b364432..b2aa294 100644 --- a/pkg/agent/service/agent.go +++ b/pkg/agent/service/agent.go @@ -482,12 +482,6 @@ func (a *Agent) collectAndSendAllMetrics(conn *safeConn, manager *collector.Mana hasError = true } - // 系统负载指标 - if err := manager.CollectAndSendLoad(conn); err != nil { - log.Printf("⚠️ 发送负载指标失败: %v", err) - hasError = true - } - // 主机信息 if err := manager.CollectAndSendHost(conn); err != nil { log.Printf("⚠️ 发送主机信息失败: %v", err) diff --git a/web/src/api/agent.ts b/web/src/api/agent.ts index 866a0c3..75767ee 100644 --- a/web/src/api/agent.ts +++ b/web/src/api/agent.ts @@ -8,8 +8,8 @@ export interface ListAgentsResponse { export interface GetAgentMetricsRequest { agentId: string; - type: 'cpu' | 'memory' | 'disk' | 'network' | 'network_connection' | 'load' | 'disk_io' | 'gpu' | 'temperature'; - range?: '1m' | '5m' | '15m' | '30m' | '1h'; + type: 'cpu' | 'memory' | 'disk' | 'network' | 'network_connection' | 'disk_io' | 'gpu' | 'temperature'; + range?: '1m' | '5m' | '15m' | '30m' | '1h' | '3h' | '6h' | '12h' | '1d' | '24h' | '3d' | '7d' | '30d'; } export interface GetAgentMetricsResponse { @@ -74,7 +74,7 @@ export const getAgentLatestMetrics = (agentId: string) => { export interface GetNetworkMetricsByInterfaceRequest { agentId: string; - range?: '1m' | '5m' | '15m' | '30m' | '1h'; + range?: '1m' | '5m' | '15m' | '30m' | '1h' | '3h' | '6h' | '12h' | '1d' | '24h' | '3d' | '7d' | '30d'; } export interface NetworkMetricByInterface { diff --git a/web/src/pages/Public/ServerDetail.tsx b/web/src/pages/Public/ServerDetail.tsx index ab25cca..4f4d2a5 100644 --- a/web/src/pages/Public/ServerDetail.tsx +++ b/web/src/pages/Public/ServerDetail.tsx @@ -94,6 +94,12 @@ const timeRangeOptions = [ {label: '15分钟', value: '15m'}, {label: '30分钟', value: '30m'}, {label: '1小时', value: '1h'}, + {label: '3小时', value: '3h'}, + {label: '6小时', value: '6h'}, + {label: '12小时', value: '12h'}, + {label: '1天', value: '1d'}, + {label: '3天', value: '3d'}, + {label: '7天', value: '7d'}, ] as const; type TimeRange = typeof timeRangeOptions[number]['value']; @@ -230,9 +236,9 @@ const TimeRangeSelector = ({ key={option.value} type="button" onClick={() => onChange(option.value)} - className={`rounded-lg border px-3 py-1.5 text-sm transition ${ + className={`rounded-lg border px-3 py-1.5 text-xs sm:text-sm font-medium transition whitespace-nowrap ${ isActive - ? 'border-blue-200 dark:border-blue-400 bg-blue-600 dark:bg-blue-500 text-white' + ? 'border-blue-200 dark:border-blue-400 bg-blue-600 dark:bg-blue-500 text-white shadow-sm' : 'border-slate-200 dark:border-slate-800 bg-white dark:bg-slate-900 text-slate-600 dark:text-slate-300 hover:border-blue-200 dark:hover:border-blue-500 hover:text-blue-600 dark:hover:text-blue-200' }`} > @@ -744,19 +750,19 @@ const ServerDetail = () => { }); cards.push({ - key: 'load', - icon: Activity, - title: '系统负载', - usagePercent: latestMetrics.load - ? `${latestMetrics.load.load1.toFixed(2)}` + key: 'network', + icon: Network, + title: '网络流量', + usagePercent: latestMetrics.network + ? `${formatBytes(latestMetrics.network.totalBytesSentRate)}/s` : '—', accent: 'amber', metrics: [ { - label: '1 / 5 / 15 分钟', - value: latestMetrics.load - ? `${latestMetrics.load.load1.toFixed(2)} / ${latestMetrics.load.load5.toFixed(2)} / ${latestMetrics.load.load15.toFixed(2)}` - : '-', + label: '上行 / 下行', + value: `${formatBytes(latestMetrics.network?.totalBytesSentRate)}/s ↑ / ${formatBytes( + latestMetrics.network?.totalBytesRecvRate, + )}/s ↓`, }, { label: '网络累计', diff --git a/web/src/pages/Settings/MetricsConfig.tsx b/web/src/pages/Settings/MetricsConfig.tsx new file mode 100644 index 0000000..829d706 --- /dev/null +++ b/web/src/pages/Settings/MetricsConfig.tsx @@ -0,0 +1,288 @@ +import {useEffect} from 'react'; +import {App, Button, Card, Form, InputNumber, Space, Spin} from 'antd'; +import {Database, Clock, BarChart3} from 'lucide-react'; +import {useMutation, useQuery, useQueryClient} from '@tanstack/react-query'; +import type {MetricsConfig} from '../../api/metrics-config'; +import {getMetricsConfig, saveMetricsConfig} from '../../api/metrics-config'; +import {getErrorMessage} from '../../lib/utils'; + +const MetricsConfigComponent = () => { + const [form] = Form.useForm(); + const {message: messageApi} = App.useApp(); + const queryClient = useQueryClient(); + + // 获取指标配置 + const {data: metricsConfig, isLoading} = useQuery({ + queryKey: ['metricsConfig'], + queryFn: getMetricsConfig, + }); + + // 保存指标配置 mutation + const saveMutation = useMutation({ + mutationFn: saveMetricsConfig, + onSuccess: () => { + messageApi.success('配置保存成功'); + queryClient.invalidateQueries({queryKey: ['metricsConfig']}); + }, + onError: (error: unknown) => { + messageApi.error(getErrorMessage(error, '保存配置失败')); + }, + }); + + // 初始化表单 + useEffect(() => { + if (metricsConfig) { + form.setFieldsValue({ + retentionHours: metricsConfig.retentionHours, + maxQueryPoints: metricsConfig.maxQueryPoints, + }); + } + }, [metricsConfig, form]); + + const handleSave = async () => { + try { + const values = await form.validateFields(); + saveMutation.mutate({ + retentionHours: values.retentionHours, + maxQueryPoints: values.maxQueryPoints, + } as MetricsConfig); + } catch (error) { + // 表单验证失败 + } + }; + + const handleReset = () => { + if (metricsConfig) { + form.setFieldsValue({ + retentionHours: metricsConfig.retentionHours, + maxQueryPoints: metricsConfig.maxQueryPoints, + }); + } + }; + + // 计算保留天数 + const getRetentionDays = (hours: number) => { + return (hours / 24).toFixed(1); + }; + + if (isLoading) { + return ( +
+ +
+ ); + } + + return ( +
+
+

+ + 指标数据配置 +

+

配置监控指标数据的保留时长和查询限制，优化存储空间和查询性能

+
+ +
+ + {/* 数据保留策略 */} + + + 数据保留策略 +
+ } + type="inner" + > + + + + + + {({getFieldValue}) => { + const hours = getFieldValue('retentionHours'); + return hours ? ( +
+ 当前设置约为 {getRetentionDays(hours)} 天 +
+ ) : null; + }} + + +
+
+
+ 💡 保留策略说明 +
+ +
+
+ +
+
+
📊 推荐设置：
+
+
+
7 天（168小时）
+
适合日常监控
+
+
+
14 天（336小时）
+
适合周期分析
+
+
+
30 天（720小时）
+
适合长期趋势
+
+
+
+
+ + + {/* 查询性能优化 */} + + + 查询性能优化 + + } + type="inner" + > + + + + +
+
+
+ ✨ 智能聚合机制 +
+
    +
  • 查询时间范围越长，系统会自动使用更大粒度的聚合数据
    • +
  • 确保返回的数据点数不超过设定的最大值，优化传输和渲染
    • +
  • 例如：查询7天数据时自动使用1小时聚合，而不是原始秒级数据
    • +
  • 聚合使用MAX（最大值）策略，确保不会遗漏峰值
    • +
+
+
+ +
+
+
+ ⚠️ 性能考量 +
+
    +
  • 较大的点数可以提供更精细的图表，但会增加查询时间和带宽消耗
    • +
  • 建议值：720 个点（适合显示24小时数据，每2分钟一个点）
    • +
  • 前端图表通常显示 300-1000 个点效果最佳，过多的点反而影响渲染性能
    • +
  • 移动端建议使用较小的值（300-500），以节省流量和提升加载速度
    • +
+
+
+ + + {/* 聚合数据说明 */} + +
+

+ 系统会在后台自动将原始数据聚合为不同粒度的数据，以支持不同时间范围的查询： +

+
+
+
+
+ 1 分钟聚合 +
+

适合查询 15分钟 - 1小时 的数据

+
+
+
+
+ 5 分钟聚合 +
+

适合查询 1小时 - 12小时 的数据

+
+
+
+
+ 1 小时聚合 +
+

适合查询 12小时 - 7天 的数据

+
+
+

+ 💡 提示：系统会自动选择最合适的聚合粒度，无需手动配置 +

+
+ + + {/* 保存按钮 */} + + + + + + + + + + ); +}; + +export default MetricsConfigComponent; diff --git a/web/src/pages/Settings/SystemConfig.tsx b/web/src/pages/Settings/SystemConfig.tsx index 542220f..5921ac2 100644 --- a/web/src/pages/Settings/SystemConfig.tsx +++ b/web/src/pages/Settings/SystemConfig.tsx @@ -20,7 +20,7 @@ const SystemConfigComponent = () => { queryFn: getSystemConfig, }); - // 保存 mutation + // 保存系统配置 mutation const saveMutation = useMutation({ mutationFn: saveSystemConfig, onSuccess: () => { @@ -34,7 +34,7 @@ const SystemConfigComponent = () => { }, }); - // 初始化表单值 + // 初始化系统配置表单 useEffect(() => { if (config) { form.setFieldsValue({ diff --git a/web/src/pages/Settings/index.tsx b/web/src/pages/Settings/index.tsx index 974b19e..0c8a3b8 100644 --- a/web/src/pages/Settings/index.tsx +++ b/web/src/pages/Settings/index.tsx @@ -1,8 +1,9 @@ import {Tabs} from 'antd'; -import {Bell, MessageSquare, Settings2} from 'lucide-react'; +import {Bell, MessageSquare, Settings2, Database} from 'lucide-react'; import AlertSettings from './AlertSettings'; import NotificationChannels from './NotificationChannels'; import SystemConfig from './SystemConfig'; +import MetricsConfig from './MetricsConfig'; import {PageHeader} from "../../components"; const Settings = () => { @@ -17,6 +18,16 @@ const Settings = () => { ), children: , }, + { + key: 'metrics', + label: ( + + + 指标数据配置 + + ), + children: , + }, { key: 'channels', label: ( diff --git a/web/src/types/index.ts b/web/src/types/index.ts index 5a70df8..6e74b30 100644 --- a/web/src/types/index.ts +++ b/web/src/types/index.ts @@ -58,15 +58,6 @@ export interface AggregatedNetworkMetric { interface: string; maxSentRate: number; maxRecvRate: number; - totalSent: number; - totalRecv: number; -} - -export interface AggregatedLoadMetric { - timestamp: number; - maxLoad1: number; - maxLoad5: number; - maxLoad15: number; } export interface AggregatedDiskMetric { @@ -140,15 +131,6 @@ export interface MemoryMetric { swapFree: number; } -export interface LoadMetric { - id: string; - agentId: string; - timestamp: number; - load1: number; - load5: number; - load15: number; -} - // 磁盘汇总数据 export interface DiskSummary { avgUsagePercent: number; // 平均使用率 @@ -405,7 +387,6 @@ export interface LatestMetrics { disk?: DiskSummary; // 改为汇总数据 network?: NetworkSummary; // 改为汇总数据 networkConnection?: NetworkConnectionMetric; // 网络连接统计 - load?: LoadMetric; host?: HostMetric; // 主机信息 gpu?: GPUMetric[]; // GPU 列表 temperature?: TemperatureMetric[]; // 温度传感器列表