telegraf/plugins/inputs/prometheus/prometheus.go

252 lines
6.1 KiB
Go
Raw Permalink Normal View History

package prometheus
import (
"errors"
"fmt"
2016-03-01 16:12:23 +00:00
"io/ioutil"
"log"
"net"
"net/http"
"net/url"
"sync"
2016-02-29 16:52:58 +00:00
"time"
"github.com/influxdata/telegraf"
"github.com/influxdata/telegraf/internal"
"github.com/influxdata/telegraf/plugins/inputs"
)
const acceptHeader = `application/vnd.google.protobuf;proto=io.prometheus.client.MetricFamily;encoding=delimited;q=0.7,text/plain;version=0.0.4;q=0.3`
type Prometheus struct {
// An array of urls to scrape metrics from.
URLs []string `toml:"urls"`
// An array of Kubernetes services to scrape metrics from.
KubernetesServices []string
// Bearer Token authorization file path
BearerToken string `toml:"bearer_token"`
ResponseTimeout internal.Duration `toml:"response_timeout"`
// Path to CA file
SSLCA string `toml:"ssl_ca"`
// Path to host cert file
SSLCert string `toml:"ssl_cert"`
// Path to cert key file
SSLKey string `toml:"ssl_key"`
// Use SSL but skip chain & host verification
InsecureSkipVerify bool
client *http.Client
}
var sampleConfig = `
## An array of urls to scrape metrics from.
urls = ["http://localhost:9100/metrics"]
## An array of Kubernetes services to scrape metrics from.
2017-09-18 23:21:45 +00:00
# kubernetes_services = ["http://my-service-dns.my-namespace:9100/metrics"]
## Use bearer token for authorization
# bearer_token = /path/to/bearer/token
## Specify timeout duration for slower prometheus clients (default is 3s)
# response_timeout = "3s"
## Optional SSL Config
# ssl_ca = /path/to/cafile
# ssl_cert = /path/to/certfile
# ssl_key = /path/to/keyfile
## Use SSL but skip chain & host verification
# insecure_skip_verify = false
2015-08-26 15:21:39 +00:00
`
func (p *Prometheus) SampleConfig() string {
return sampleConfig
}
func (p *Prometheus) Description() string {
return "Read metrics from one or many prometheus clients"
}
var ErrProtocolError = errors.New("prometheus protocol error")
func (p *Prometheus) AddressToURL(u *url.URL, address string) *url.URL {
host := address
if u.Port() != "" {
host = address + ":" + u.Port()
}
reconstructedURL := &url.URL{
Scheme: u.Scheme,
Opaque: u.Opaque,
User: u.User,
Path: u.Path,
RawPath: u.RawPath,
ForceQuery: u.ForceQuery,
RawQuery: u.RawQuery,
Fragment: u.Fragment,
Host: host,
}
return reconstructedURL
}
type URLAndAddress struct {
OriginalURL *url.URL
URL *url.URL
Address string
}
func (p *Prometheus) GetAllURLs() ([]URLAndAddress, error) {
allURLs := make([]URLAndAddress, 0)
for _, u := range p.URLs {
URL, err := url.Parse(u)
if err != nil {
log.Printf("prometheus: Could not parse %s, skipping it. Error: %s", u, err)
continue
}
allURLs = append(allURLs, URLAndAddress{URL: URL, OriginalURL: URL})
}
for _, service := range p.KubernetesServices {
URL, err := url.Parse(service)
if err != nil {
return nil, err
}
resolvedAddresses, err := net.LookupHost(URL.Hostname())
if err != nil {
log.Printf("prometheus: Could not resolve %s, skipping it. Error: %s", URL.Host, err)
continue
}
for _, resolved := range resolvedAddresses {
serviceURL := p.AddressToURL(URL, resolved)
allURLs = append(allURLs, URLAndAddress{URL: serviceURL, Address: resolved, OriginalURL: URL})
}
}
return allURLs, nil
}
// Reads stats from all configured servers accumulates stats.
// Returns one of the errors encountered while gather stats (if any).
func (p *Prometheus) Gather(acc telegraf.Accumulator) error {
if p.client == nil {
client, err := p.createHttpClient()
if err != nil {
return err
}
p.client = client
}
var wg sync.WaitGroup
allURLs, err := p.GetAllURLs()
if err != nil {
return err
}
for _, URL := range allURLs {
wg.Add(1)
go func(serviceURL URLAndAddress) {
defer wg.Done()
acc.AddError(p.gatherURL(serviceURL, acc))
}(URL)
}
wg.Wait()
2017-04-24 18:13:26 +00:00
return nil
}
2016-02-29 16:52:58 +00:00
var tr = &http.Transport{
ResponseHeaderTimeout: time.Duration(3 * time.Second),
}
var client = &http.Client{
Transport: tr,
Timeout: time.Duration(4 * time.Second),
}
func (p *Prometheus) createHttpClient() (*http.Client, error) {
tlsCfg, err := internal.GetTLSConfig(
p.SSLCert, p.SSLKey, p.SSLCA, p.InsecureSkipVerify)
if err != nil {
return nil, err
}
client := &http.Client{
Transport: &http.Transport{
TLSClientConfig: tlsCfg,
DisableKeepAlives: true,
},
Timeout: p.ResponseTimeout.Duration,
}
return client, nil
}
func (p *Prometheus) gatherURL(u URLAndAddress, acc telegraf.Accumulator) error {
var req, err = http.NewRequest("GET", u.URL.String(), nil)
req.Header.Add("Accept", acceptHeader)
var token []byte
var resp *http.Response
if p.BearerToken != "" {
token, err = ioutil.ReadFile(p.BearerToken)
if err != nil {
return err
}
req.Header.Set("Authorization", "Bearer "+string(token))
}
resp, err = p.client.Do(req)
if err != nil {
return fmt.Errorf("error making HTTP request to %s: %s", u.URL, err)
}
defer resp.Body.Close()
if resp.StatusCode != http.StatusOK {
return fmt.Errorf("%s returned HTTP status %s", u.URL, resp.Status)
}
2016-03-01 16:12:23 +00:00
body, err := ioutil.ReadAll(resp.Body)
if err != nil {
return fmt.Errorf("error reading body: %s", err)
}
metrics, err := Parse(body, resp.Header)
2016-03-01 16:12:23 +00:00
if err != nil {
return fmt.Errorf("error reading metrics for %s: %s",
u.URL, err)
2016-03-01 16:12:23 +00:00
}
// Add (or not) collected metrics
for _, metric := range metrics {
tags := metric.Tags()
// strip user and password from URL
u.OriginalURL.User = nil
tags["url"] = u.OriginalURL.String()
if u.Address != "" {
tags["address"] = u.Address
}
switch metric.Type() {
case telegraf.Counter:
acc.AddCounter(metric.Name(), metric.Fields(), tags, metric.Time())
case telegraf.Gauge:
acc.AddGauge(metric.Name(), metric.Fields(), tags, metric.Time())
case telegraf.Summary:
acc.AddSummary(metric.Name(), metric.Fields(), tags, metric.Time())
case telegraf.Histogram:
acc.AddHistogram(metric.Name(), metric.Fields(), tags, metric.Time())
default:
acc.AddFields(metric.Name(), metric.Fields(), tags, metric.Time())
}
}
return nil
}
func init() {
inputs.Add("prometheus", func() telegraf.Input {
2016-12-17 13:10:33 +00:00
return &Prometheus{ResponseTimeout: internal.Duration{Duration: time.Second * 3}}
})
}