Exemplo n.º 1
0
func UpdateStorageProfileUtilizationToSummaries(ctxt string, cluster models.Cluster) {
	pluginIndex := monitoring.GetPluginIndex(monitoring.STORAGE_PROFILE_UTILIZATION, cluster.Monitoring.Plugins)
	if pluginIndex != -1 {
		if storageProfileUtilization, err := ComputeStorageProfileUtilization(
			bson.M{"clusterid": cluster.ClusterId},
			cluster.Monitoring.Plugins[pluginIndex].Configs); err != nil {
			logger.Get().Error("%s - Failed to fetch storage profile utilization of cluster %v.Error %v", ctxt, cluster.Name, err)
		} else {
			UpdateDb(bson.M{"clusterid": cluster.ClusterId}, bson.M{"storageprofileusage": storageProfileUtilization}, models.COLL_NAME_CLUSTER_SUMMARY, ctxt)
		}
	}

	systemthresholds := monitoring.GetSystemDefaultThresholdValues()
	utilization, err := ComputeStorageProfileUtilization(nil, systemthresholds[monitoring.STORAGE_PROFILE_UTILIZATION].Configs)
	if err != nil {
		logger.Get().Error("%s - Error updating the storage profile utilization to system summary. Error %v", ctxt, err)
	} else {
		UpdateDb(bson.M{"name": monitoring.SYSTEM}, bson.M{"storageprofileusage": utilization}, models.COLL_NAME_SKYRING_UTILIZATION, ctxt)
	}
}
Exemplo n.º 2
0
func AnalyseThresholdBreach(ctxt string, utilizationType string, resourceName string, resourceUtilization float64, cluster models.Cluster) (models.Event, bool, error) {
	var event models.Event
	timeStamp := time.Now()
	pluginIndex := monitoring.GetPluginIndex(utilizationType, cluster.Monitoring.Plugins)
	if pluginIndex == -1 {
		logger.Get().Error(
			"%s - The cluster %s is not configured to monitor threshold breaches for %v and hence cannot monitor utilization of %v",
			ctxt, cluster.Name, utilizationType, resourceName)
		return models.Event{}, false,
			fmt.Errorf(
				"%s - The cluster %s is not configured to monitor threshold breaches for %v and hence cannot monitor utilization of %v",
				ctxt, cluster.Name, utilizationType, resourceName)
	}
	plugin := cluster.Monitoring.Plugins[pluginIndex]
	var applicableConfig monitoring.PluginConfig
	var applicableThresholdValue float64
	var message string

	var entityIdentifier string
	entityId, entityIdFetchError := getEntityIdFromNameAndUtilizationType(utilizationType, resourceName, cluster)
	if entityIdFetchError != nil {
		logger.Get().Error("%s - Error fetching the id for %v in cluster %v",
			ctxt, resourceName, cluster.Name)
		return models.Event{}, false, fmt.Errorf("%s - Error fetching the id for %v in cluster %v",
			ctxt, resourceName, cluster.Name)
	}
	entityIdentifier = (*entityId).String()

	if len(plugin.Configs) == 0 {
		return models.Event{}, false, fmt.Errorf("%s - No threshold configurations found for %v of %v in cluster %v",
			ctxt, utilizationType, resourceName, cluster.Name)
	}

	for _, config := range plugin.Configs {
		if config.Category == monitoring.THRESHOLD {
			currentThresholdValue, thresholdValError := strconv.ParseFloat(config.Value, 64)
			if thresholdValError != nil {
				logger.Get().Error("%s - %v-%v configuration for %v in cluster %v could not be converted.Err %v\n",
					ctxt, config.Category, config.Type, plugin.Name, cluster.Name, thresholdValError)
			}

			if resourceUtilization > currentThresholdValue && applicableThresholdValue < currentThresholdValue {
				applicableConfig = config
				applicableThresholdValue = currentThresholdValue
			}
		}
	}

	if applicableConfig.Type == "" {
		applicableConfig.Type = models.STATUS_OK
		message = fmt.Sprintf("%v of cluster %v with value %v back to normal",
			plugin.Name, cluster.Name, resourceUtilization)
	}

	tEvent := models.ThresholdEvent{
		ClusterId:         cluster.ClusterId,
		UtilizationType:   plugin.Name,
		ThresholdSeverity: strings.ToUpper(applicableConfig.Type),
		TimeStamp:         timeStamp,
		EntityId:          *entityId,
		EntityName:        resourceName,
	}

	err, isRaiseEvent := UpdateThresholdInfoToTable(tEvent)
	if err != nil {
		logger.Get().Error("%s - %v", ctxt, err)
	}

	if isRaiseEvent {
		event = models.Event{
			Timestamp: timeStamp,
			ClusterId: cluster.ClusterId,
			NodeId:    cluster.ClusterId,
			Tag: fmt.Sprintf("skyring/%v/cluster/%v/threshold/%v/%v",
				cluster.Type, cluster.ClusterId, plugin.Name, strings.ToUpper(applicableConfig.Type)),
			Tags: map[string]string{
				models.CURRENT_VALUE:   strconv.FormatFloat(resourceUtilization, 'E', -1, 64),
				models.ENTITY_ID:       entityIdentifier,
				models.PLUGIN:          plugin.Name,
				models.THRESHOLD_TYPE:  strings.ToUpper(applicableConfig.Type),
				models.THRESHOLD_VALUE: strconv.FormatFloat(applicableThresholdValue, 'E', -1, 64),
				models.ENTITY_NAME:     resourceName,
			},
			Severity: strings.ToUpper(applicableConfig.Type),
		}
		if message == "" {
			message = fmt.Sprintf("%v of cluster %v with value %v breached %v threshold of value %v",
				plugin.Name, cluster.Name, resourceUtilization, strings.ToUpper(applicableConfig.Type), applicableThresholdValue)
		}
		event.Message = message
		return event, true, nil
	}
	return models.Event{}, false, nil
}