Replace WMI query to get the system/diskio metrics for Windows

CHANGELOG.next.asciidoc

@@ -91,25 +91,6 @@ https://github.com/elastic/beats/compare/v7.0.0-alpha2...master[Check the HEAD d
 - Add _bucket to histogram metrics in Prometheus Collector {pull}11578[11578]
 - Prevent the docker/memory metricset from processing invalid events before container start {pull}11676[11676]
 - Change `add_cloud_metadata` processor to not overwrite `cloud` field when it already exist in the event. {pull}11612[11612] {issue}11305[11305]
-- Fix panics in vsphere module when certain values where not returned by the API. {pull}9784[9784]
-- Fix pod UID metadata enrichment in Kubernetes module. {pull}10081[10081]
-- Fix issue that would prevent collection of processes without command line on Windows. {pull}10196[10196]
-- Fixed data type for tags field in `docker/container` metricset {pull}10307[10307]
-- Fixed data type for tags field in `docker/image` metricset {pull}10307[10307]
-- Fixed data type for isr field in `kafka/partition` metricset {pull}10307[10307]
-- Fixed data types for various hosts fields in `mongodb/replstatus` metricset {pull}10307[10307]
-- Added function to close sql database connection. {pull}10355[10355]
-- Fix issue with `elasticsearch/node_stats` metricset (x-pack) not indexing `source_node` field. {pull}10639[10639]
-- Migrate docker autodiscover to ECS. {issue}10757[10757] {pull}10862[10862]
-- Fix issue in kubernetes module preventing usage percentages to be properly calculated. {pull}10946[10946]
-- Fix for not reusable http client leading to connection leaks in Jolokia module {pull}11014[11014]
-- Fix parsing error using GET in Jolokia module. {pull}11075[11075] {issue}11071[11071]
-- Collect metrics when EC2 instances are not in running state. {issue}11008[11008] {pull}11023[11023]
-- Change ECS field cloud.provider to aws. {pull}11023[11023]
-- Add documentation about jolokia autodiscover fields. {issue}10925[10925] {pull}10979[10979]
-- Add missing aws.ec2.instance.state.name into fields.yml. {issue}11219[11219] {pull}11221[11221]
-- Fix ec2 metricset to collect metrics from Cloudwatch with the same timestamp. {pull}11142[11142]
-- Fix potential memory leak in stopped docker metricsets {pull}11294[11294]
 - Change diskio metrics retrieval method (only for Windows) from wmi query to DeviceIOControl function using the IOCTL_DISK_PERFORMANCE control code {pull}11635[11635]
 
 *Packetbeat*

metricbeat/module/system/diskio/diskstat_windows_helper.go

@@ -23,6 +23,8 @@ import (
 	"syscall"
 	"unsafe"
 
+	"github.com/elastic/beats/libbeat/logp"
+
 	"github.com/pkg/errors"
 	"golang.org/x/sys/windows/registry"
 
@@ -39,6 +41,7 @@ var (
 	modkernel32                 = syscall.NewLazyDLL("kernel32.dll")
 	procGetLogicalDriveStringsW = modkernel32.NewProc("GetLogicalDriveStringsW")
 	procGetDriveTypeW           = modkernel32.NewProc("GetDriveTypeW")
+	logger                      = logp.NewLogger("diskio")
 )
 
 type logicalDrive struct {
@@ -98,7 +101,7 @@ func ioCounters(names ...string) (map[string]disk.IOCountersStat, error) {
 	return ret, nil
 }
 
-// i0Counter calls syscal func CreateFile to generate a handler then executes the DeviceIoControl func in order to retrieve the metrics.
+// ioCounter calls syscal func CreateFile to generate a handler then executes the DeviceIoControl func in order to retrieve the metrics.
 func ioCounter(path string) (diskPerformance, error) {
 	var diskPerformance diskPerformance
 	var diskPerformanceSize uint32
@@ -142,6 +145,7 @@ func enablePerformanceCounters() error {
 	if err = key.SetDWordValue("EnableCounterForIoctl", 1); err != nil {
 		return errors.Errorf("cannot create EnableCounterForIoctl key in the registry in order to enable the performance counters: %s", err)
 	}
+	logger.Info("The key with the name EnableCounterForIoctl has been added in the windows registry in order to enable the performance counters")
 	return nil
 }