bsc/metrics/influxdb/influxdbv2.go
dylanhuang bf68be2bd9
dep: bump the version of several important library (#1274)
* dep: upgrade secp256k1 to use btcec/v2 v2.3.2 and update insecurity pkg

* build ci: upgrade go to 1.19 and golangci-lint to 1.50.1

* docs: fix format that does not follow the goimports

* dep: redirect github.com/bnb-chain/tendermint to v0.31.13

* ci: disable GOPROXY
2023-02-02 19:36:37 +08:00

223 lines
5.6 KiB
Go

// The go-ethereum library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
package influxdb
import (
"context"
"fmt"
"time"
"github.com/ethereum/go-ethereum/log"
"github.com/ethereum/go-ethereum/metrics"
influxdb2 "github.com/influxdata/influxdb-client-go/v2"
"github.com/influxdata/influxdb-client-go/v2/api"
)
type v2Reporter struct {
reg metrics.Registry
interval time.Duration
endpoint string
token string
bucket string
organization string
namespace string
tags map[string]string
client influxdb2.Client
write api.WriteAPI
cache map[string]int64
}
// InfluxDBWithTags starts a InfluxDB reporter which will post the from the given metrics.Registry at each d interval with the specified tags
func InfluxDBV2WithTags(r metrics.Registry, d time.Duration, endpoint string, token string, bucket string, organization string, namespace string, tags map[string]string) {
rep := &v2Reporter{
reg: r,
interval: d,
endpoint: endpoint,
token: token,
bucket: bucket,
organization: organization,
namespace: namespace,
tags: tags,
cache: make(map[string]int64),
}
rep.client = influxdb2.NewClient(rep.endpoint, rep.token)
defer rep.client.Close()
// async write client
rep.write = rep.client.WriteAPI(rep.organization, rep.bucket)
errorsCh := rep.write.Errors()
// have to handle write errors in a separate goroutine like this b/c the channel is unbuffered and will block writes if not read
go func() {
for err := range errorsCh {
log.Warn("write error", "err", err.Error())
}
}()
rep.run()
}
func (r *v2Reporter) run() {
intervalTicker := time.Tick(r.interval)
pingTicker := time.Tick(time.Second * 5)
for {
select {
case <-intervalTicker:
r.send()
case <-pingTicker:
_, err := r.client.Health(context.Background())
if err != nil {
log.Warn("Got error from influxdb client health check", "err", err.Error())
}
}
}
}
func (r *v2Reporter) send() {
r.reg.Each(func(name string, i interface{}) {
now := time.Now()
namespace := r.namespace
switch metric := i.(type) {
case metrics.Counter:
v := metric.Count()
l := r.cache[name]
measurement := fmt.Sprintf("%s%s.count", namespace, name)
fields := map[string]interface{}{
"value": v - l,
}
pt := influxdb2.NewPoint(measurement, r.tags, fields, now)
r.write.WritePoint(pt)
r.cache[name] = v
case metrics.Gauge:
ms := metric.Snapshot()
measurement := fmt.Sprintf("%s%s.gauge", namespace, name)
fields := map[string]interface{}{
"value": ms.Value(),
}
pt := influxdb2.NewPoint(measurement, r.tags, fields, now)
r.write.WritePoint(pt)
case metrics.GaugeFloat64:
ms := metric.Snapshot()
measurement := fmt.Sprintf("%s%s.gauge", namespace, name)
fields := map[string]interface{}{
"value": ms.Value(),
}
pt := influxdb2.NewPoint(measurement, r.tags, fields, now)
r.write.WritePoint(pt)
case metrics.Histogram:
ms := metric.Snapshot()
if ms.Count() > 0 {
ps := ms.Percentiles([]float64{0.5, 0.75, 0.95, 0.99, 0.999, 0.9999})
measurement := fmt.Sprintf("%s%s.histogram", namespace, name)
fields := map[string]interface{}{
"count": ms.Count(),
"max": ms.Max(),
"mean": ms.Mean(),
"min": ms.Min(),
"stddev": ms.StdDev(),
"variance": ms.Variance(),
"p50": ps[0],
"p75": ps[1],
"p95": ps[2],
"p99": ps[3],
"p999": ps[4],
"p9999": ps[5],
}
pt := influxdb2.NewPoint(measurement, r.tags, fields, now)
r.write.WritePoint(pt)
}
case metrics.Meter:
ms := metric.Snapshot()
measurement := fmt.Sprintf("%s%s.meter", namespace, name)
fields := map[string]interface{}{
"count": ms.Count(),
"m1": ms.Rate1(),
"m5": ms.Rate5(),
"m15": ms.Rate15(),
"mean": ms.RateMean(),
}
pt := influxdb2.NewPoint(measurement, r.tags, fields, now)
r.write.WritePoint(pt)
case metrics.Timer:
ms := metric.Snapshot()
ps := ms.Percentiles([]float64{0.5, 0.75, 0.95, 0.99, 0.999, 0.9999})
measurement := fmt.Sprintf("%s%s.timer", namespace, name)
fields := map[string]interface{}{
"count": ms.Count(),
"max": ms.Max(),
"mean": ms.Mean(),
"min": ms.Min(),
"stddev": ms.StdDev(),
"variance": ms.Variance(),
"p50": ps[0],
"p75": ps[1],
"p95": ps[2],
"p99": ps[3],
"p999": ps[4],
"p9999": ps[5],
"m1": ms.Rate1(),
"m5": ms.Rate5(),
"m15": ms.Rate15(),
"meanrate": ms.RateMean(),
}
pt := influxdb2.NewPoint(measurement, r.tags, fields, now)
r.write.WritePoint(pt)
case metrics.ResettingTimer:
t := metric.Snapshot()
if len(t.Values()) > 0 {
ps := t.Percentiles([]float64{50, 95, 99})
val := t.Values()
measurement := fmt.Sprintf("%s%s.span", namespace, name)
fields := map[string]interface{}{
"count": len(val),
"max": val[len(val)-1],
"mean": t.Mean(),
"min": val[0],
"p50": ps[0],
"p95": ps[1],
"p99": ps[2],
}
pt := influxdb2.NewPoint(measurement, r.tags, fields, now)
r.write.WritePoint(pt)
}
}
})
// Force all unwritten data to be sent
r.write.Flush()
}