Read config values from environment for max_conns tuning

- max_conns / idle / per host are now read from env-vars and have
defaults set to 1024 for both values
- logging / metrics are collected in the client transaction
rather than via defer (this may impact throughput)
- function cache moved to use RWMutex to try to improve latency
around locking when updating cache
- logging message added to show latency in running GetReplicas
because this was observed to increase in a linear fashion under
high concurrency
- changes tested against 3-node bare-metal 1.13 K8s cluster
with kubeadm

Signed-off-by: Alex Ellis (VMware) <alexellis2@gmail.com>

gateway/plugin/external.go

@@ -6,6 +6,8 @@ package plugin
 import (
 	"bytes"
 	"encoding/json"
+	"fmt"
+	"io/ioutil"
 	"log"
 	"net"
 	"net/http"
@@ -13,10 +15,6 @@ import (
 	"strconv"
 	"time"
 
-	"fmt"
-
-	"io/ioutil"
-
 	"github.com/openfaas/faas-provider/auth"
 	"github.com/openfaas/faas/gateway/requests"
 	"github.com/openfaas/faas/gateway/scaling"
@@ -62,6 +60,8 @@ type ScaleServiceRequest struct {
 
 // GetReplicas replica count for function
 func (s ExternalServiceQuery) GetReplicas(serviceName string) (scaling.ServiceQueryResponse, error) {
+	start := time.Now()
+
 	var err error
 	var emptyServiceQueryResponse scaling.ServiceQueryResponse
 
@@ -92,6 +92,7 @@ func (s ExternalServiceQuery) GetReplicas(serviceName string) (scaling.ServiceQu
 				log.Println(urlPath, err)
 			}
 		} else {
+			log.Printf("GetReplicas took: %fs", time.Since(start).Seconds())
 			return emptyServiceQueryResponse, fmt.Errorf("server returned non-200 status code (%d) for function, %s", res.StatusCode, serviceName)
 		}
 	}
@@ -115,6 +116,8 @@ func (s ExternalServiceQuery) GetReplicas(serviceName string) (scaling.ServiceQu
 		}
 	}
 
+	log.Printf("GetReplicas took: %fs", time.Since(start).Seconds())
+
 	return scaling.ServiceQueryResponse{
 		Replicas:          function.Replicas,
 		MaxReplicas:       maxReplicas,