README.md

pipes

Pipe your metric to elasticsearch from your terminal.
use case example:

load avg a linux machine

cat /proc/loadavg | awk '{print $1}' | ./pipes -i load-avg-load-index -h $host -v

will produce

{
    "value": "1.2",
    "@timestamp": "1497399556",
    "hostname": "my-host",
}

/home directory size for each user

du -shBK  * | awk -v OFS='\t' '{print $2, $1}' | ./pipes -m key_value -i home-dir-index -h $host -v

will create 3 event, like this:

{
    "key": "joe",
    "value": "8000K",
    "@timestamp": "1497399556",
    "hostname": "my-host",
}
--
{
    "key": "ubuntu",
    "value": "16000K",
    "@timestamp": "1497399556",
    "hostname": "my-host",
}
--
{
    "key": "mark",
    "value": "32000K",
    "@timestamp": "1497399556",
    "hostname": "my-host",
}

List of all options

Usage of ./pipes:
  --dry-run
    	Enable dry-run mode (just output json, without make request)
  -h, --host string
    	Elasticsearch host (default "localhost")
  -i, --index string
    	Index name
  -m, --mode string
    	Mode (single_value|key_value) (default "single_value")
  -p, --password string
    	Basic auth password
  -P, --port string
    	Elasticsearch port (default "9200")
  -t, --type string
    	Index log type (default "log")
  -u, --user string
    	Basic auth username
  -v, --verbose
    	Verbose output