g_met's Introduction

XMet

XMet is to solve the following pains:

Various metric solutions bring the hard problems of maintaining. Even, the whole group uses unified data aggregation and demonstration solution such as: ELK (ElasticSearch LogStach Kibana) or EFK (ElasticSearch FluentD Kibana). Because, the way to report metric log/data and the format of log/data are normally different.
The inefficient metric data creating and extracting ways take huge CPU resource.
So many different configurations (fluentD configurations, elasticsearch mappings), which are hard to maintain.
Creating metric monitor is a time-consuming task, which needs to write and debug the configurations for the related systems (fluentD, elastic search)

XMet is to help you create the program running metrics quickly and easily.

Easy to use. To send the metrics, just one line code is needed.
Easy to extend. GMet is based on the plugin structure, it is open to extend and by ioslated with the abstract layer, the changes are transparent to the clients.
Easy to maintain. Use the unified API to avoid varied and messy implementations. Deal with different strategies on log and metric data by separating them.

GMet is the golang API of XMet.

//The following example is to demonstrate how to use GMet.
func main() {
	//create GMet instance by given default writer and the formatter
	gmet := CreateGMetInstanceByDefault("../configs/g_met_seelog/g_met_log.xml")
	for i := 0; i < 100; i++ {
		gmet.Send(Metric("input_bytes", rand.Intn(100)),
			Metric("output_bytes", rand.Intn(100)))
		gmet.Flush() //in your real case, DON'T flush for each sending.
		//For seelog writer, the auto-flushing can be set in the log configuration
		time.Sleep(time.Second * 1)

	}
}