Support custom test configurations

test/integration/ipamperf/BUILD

@@ -60,6 +60,7 @@ go_library(
         "//vendor/google.golang.org/api/compute/v0.beta:go_default_library",
         "//vendor/google.golang.org/api/compute/v1:go_default_library",
         "//vendor/k8s.io/api/core/v1:go_default_library",
+        "//vendor/k8s.io/apimachinery/pkg/api/errors:go_default_library",
         "//vendor/k8s.io/apimachinery/pkg/api/resource:go_default_library",
         "//vendor/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
         "//vendor/k8s.io/client-go/informers:go_default_library",

test/integration/ipamperf/README.md

@@ -29,10 +29,18 @@ The runner scripts support a few different options:
 ```shell
 ./test-performance.sh -h
 usage: ./test-performance.sh [-h] [-d] [-r <pattern>] [-o <filename>]
+usage: ./test-performance.sh <options>
  -h display this help message
  -d enable debug logs in tests
  -r <pattern> regex pattern to match for tests
  -o <filename> file to write JSON formatted results to
+ -p <id> enable cpu and memory profiles, output written to mem-<id>.out and cpu-<id>.out
+ -c enable custom test configuration
+ -a <name> allocator name, one of RangeAllocator, CloudAllocator, IPAMFromCluster, IPAMFromCloud
+ -k <num> api server qps for allocator
+ -n <num> number of nodes to simulate
+ -m <num> api server qps for node creation
+ -l <num> gce cloud endpoint qps
 ```
 
 The tests follow the pattern TestPerformance/{AllocatorType}-KubeQPS{X}-Nodes{Y}, where AllocatorType 
@@ -56,6 +64,19 @@ So to run the test for CloudAllocator with 10 nodes, one can run
 At the end of the test, a JSON format of the results for all the tests run is printed. Passing the -o option 
 allows for also saving this JSON to a named file.
 
+### Profiling the code
+It's possible to get the CPU and memory profiles of code during test execution by using the ```-p``` option.
+The CPU and memory profiles are generated in the same directory with the file names set to ```cpu-<id>.out```
+and ```cpu-<id>.out```, where ```<id>``` is the argument value. Typicall pattern is to put in the number
+of nodes being simulated as the id, or 'all' in case running the full suite.
+
+### Custom Test Configuration
+It's also possible to run a custom test configuration by passing the -c option. With this option, it then
+possible to specify the number of nodes to simulate and the API server qps values for creation, 
+IPAM allocation and cloud endpoint, along with the allocator name to run. The defaults values for the 
+qps parmeters are 30 for IPAM allocation, 100 for node creation and 30 for the cloud endpoint, and the
+default allocator is the RangeAllocator.
+
 Code Organization
 -----
 The core of the tests are defined in [ipam_test.go](ipam_test.go), using the t.Run() helper to control parallelism

test/integration/ipamperf/ipam_test.go

@@ -126,9 +126,13 @@ func TestPerformance(t *testing.T) {
 		tests      []*Config
 	)
 
-	for _, numNodes := range []int{10, 100} {
-		for _, alloc := range []ipam.CIDRAllocatorType{ipam.RangeAllocatorType, ipam.CloudAllocatorType, ipam.IPAMFromClusterAllocatorType, ipam.IPAMFromCloudAllocatorType} {
-			tests = append(tests, &Config{AllocatorType: alloc, NumNodes: numNodes, CreateQPS: numNodes, KubeQPS: 10, CloudQPS: 10})
+	if isCustom {
+		tests = append(tests, customConfig)
+	} else {
+		for _, numNodes := range []int{10, 100} {
+			for _, alloc := range []ipam.CIDRAllocatorType{ipam.RangeAllocatorType, ipam.CloudAllocatorType, ipam.IPAMFromClusterAllocatorType, ipam.IPAMFromCloudAllocatorType} {
+				tests = append(tests, &Config{AllocatorType: alloc, NumNodes: numNodes, CreateQPS: numNodes, KubeQPS: 10, CloudQPS: 10})
+			}
 		}
 	}
 

test/integration/ipamperf/main_test.go

@@ -20,13 +20,47 @@ import (
 	"flag"
 	"testing"
 
+	"github.com/golang/glog"
+	"k8s.io/kubernetes/pkg/controller/nodeipam/ipam"
 	"k8s.io/kubernetes/test/integration/framework"
 )
 
-var resultsLogFile string
+var (
+	resultsLogFile string
+	isCustom       bool
+	customConfig   = &Config{
+		NumNodes:      10,
+		KubeQPS:       30,
+		CloudQPS:      30,
+		CreateQPS:     100,
+		AllocatorType: ipam.RangeAllocatorType,
+	}
+)
 
 func TestMain(m *testing.M) {
+	allocator := string(ipam.RangeAllocatorType)
+
 	flag.StringVar(&resultsLogFile, "log", "", "log file to write JSON results to")
+	flag.BoolVar(&isCustom, "custom", false, "enable custom test configuration")
+	flag.StringVar(&allocator, "allocator", allocator, "allocator to use")
+	flag.IntVar(&customConfig.KubeQPS, "kube-qps", customConfig.KubeQPS, "API server qps for allocations")
+	flag.IntVar(&customConfig.NumNodes, "num-nodes", 10, "number of nodes to simulate")
+	flag.IntVar(&customConfig.CreateQPS, "create-qps", customConfig.CreateQPS, "API server qps for node creation")
+	flag.IntVar(&customConfig.CloudQPS, "cloud-qps", customConfig.CloudQPS, "GCE Cloud qps limit")
 	flag.Parse()
+
+	switch allocator {
+	case string(ipam.RangeAllocatorType):
+		customConfig.AllocatorType = ipam.RangeAllocatorType
+	case string(ipam.CloudAllocatorType):
+		customConfig.AllocatorType = ipam.CloudAllocatorType
+	case string(ipam.IPAMFromCloudAllocatorType):
+		customConfig.AllocatorType = ipam.IPAMFromCloudAllocatorType
+	case string(ipam.IPAMFromClusterAllocatorType):
+		customConfig.AllocatorType = ipam.IPAMFromClusterAllocatorType
+	default:
+		glog.Fatalf("Unknown allocator type: %s", allocator)
+	}
+
 	framework.EtcdMain(m.Run)
 }

test/integration/ipamperf/results.go

@@ -95,7 +95,7 @@ func NewObserver(clientSet *clientset.Clientset, numNodes int) *Observer {
 // StartObserving starts an asynchronous loop to monitor for node changes.
 // Call Results() to get the test results after starting observer.
 func (o *Observer) StartObserving() error {
-	go o.monitor()
+	o.monitor()
 	glog.Infof("Test observer started")
 	return nil
 }
@@ -173,6 +173,9 @@ func (o *Observer) monitor() {
 				nTime.allocated = time.Now()
 				nTime.podCIDR = newNode.Spec.PodCIDR
 				o.numAllocated++
+				if o.numAllocated%10 == 0 {
+					glog.Infof("progress: %d/%d - %.2d%%", o.numAllocated, o.numNodes, (o.numAllocated * 100.0 / o.numNodes))
+				}
 				// do following check only if numAllocated is modified, as otherwise, redundant updates
 				// can cause wg.Done() to be called multiple times, causing a panic
 				if o.numAdded == o.numNodes && o.numAllocated == o.numNodes {

test/integration/ipamperf/test-performance.sh

@@ -20,17 +20,25 @@ set -o pipefail
 
 TEST_ARGS=""
 RUN_PATTERN=".*"
+PROFILE_OPTS=""
 
 function usage() {
-  echo "usage: $0 [-h] [-d] [-r <pattern>] [-o <filename>]"
+  echo "usage: $0 <options>"
   echo " -h display this help message"
   echo " -d enable debug logs in tests"
   echo " -r <pattern> regex pattern to match for tests"
   echo " -o <filename> file to write JSON formatted results to"
+  echo " -p <id> enable cpu and memory profiles, output written to mem-<id>.out and cpu-<id>.out"
+  echo " -c enable custom test configuration"
+  echo " -a <name> allocator name, one of RangeAllocator, CloudAllocator, IPAMFromCluster, IPAMFromCloud"
+  echo " -k <num> api server qps for allocator"
+  echo " -n <num> number of nodes to simulate"
+  echo " -m <num> api server qps for node creation"
+  echo " -l <num> gce cloud endpoint qps"
   exit 1
 }
 
-while getopts ":hdr:o:" opt; do
+while getopts ":hdr:o:p:ca:k:n:m:l:" opt; do
   case ${opt} in
     d) TEST_ARGS="${TEST_ARGS} -v=6"
       ;;
@@ -38,9 +46,23 @@ while getopts ":hdr:o:" opt; do
       ;;
     o) TEST_ARGS="${TEST_ARGS} -log ${OPTARG}"
       ;;
-    h) ::usage
+    p) PROFILE_OPTS="-memprofile mem-${OPTARG}.out -cpuprofile cpu-${OPTARG}.out"
       ;;
-    \?) ::usage
+    c) TEST_ARGS="${TEST_ARGS} -custom"
+      ;;
+    a) TEST_ARGS="${TEST_ARGS} -allocator ${OPTARG}"
+      ;;
+    k) TEST_ARGS="${TEST_ARGS} -kube-qps ${OPTARG}"
+      ;;
+    n) TEST_ARGS="${TEST_ARGS} -num-nodes ${OPTARG}"
+      ;;
+    m) TEST_ARGS="${TEST_ARGS} -create-qps ${OPTARG}"
+      ;;
+    l) TEST_ARGS="${TEST_ARGS} -cloud-qps ${OPTARG}"
+      ;;
+    h) usage
+      ;;
+    \?) usage
       ;;
   esac
 done
@@ -65,5 +87,5 @@ kube::etcd::start
 
 # Running IPAM tests. It might take a long time.
 kube::log::status "performance test (IPAM) start"
-go test -test.run=${RUN_PATTERN} -test.timeout=60m -test.short=false -v -args ${TEST_ARGS}
+go test ${PROFILE_OPTS} -test.run=${RUN_PATTERN} -test.timeout=60m -test.short=false -v -args ${TEST_ARGS}
 kube::log::status "... IPAM tests finished."

test/integration/ipamperf/util.go

@@ -17,8 +17,11 @@ limitations under the License.
 package ipamperf
 
 import (
+	"time"
+
 	"github.com/golang/glog"
 	"k8s.io/api/core/v1"
+	"k8s.io/apimachinery/pkg/api/errors"
 	"k8s.io/apimachinery/pkg/api/resource"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 	clientset "k8s.io/client-go/kubernetes"
@@ -26,6 +29,11 @@ import (
 	"k8s.io/kubernetes/pkg/api/testapi"
 )
 
+const (
+	maxCreateRetries = 10
+	retryDelay       = 10 * time.Second
+)
+
 var (
 	baseNodeTemplate = &v1.Node{
 		ObjectMeta: metav1.ObjectMeta{
@@ -71,7 +79,17 @@ func createNodes(apiURL string, config *Config) error {
 	})
 	glog.Infof("Creating %d nodes", config.NumNodes)
 	for i := 0; i < config.NumNodes; i++ {
-		if _, err := clientSet.CoreV1().Nodes().Create(baseNodeTemplate); err != nil {
+		var err error
+		for j := 0; j < maxCreateRetries; j++ {
+			if _, err = clientSet.CoreV1().Nodes().Create(baseNodeTemplate); err != nil && errors.IsServerTimeout(err) {
+				glog.Infof("Server timeout creating nodes, retrying after %v", retryDelay)
+				time.Sleep(retryDelay)
+				continue
+			}
+			break
+		}
+		if err != nil {
+			glog.Errorf("Error creating nodes: %v", err)
 			return err
 		}
 	}