Merge pull request #2036 from NetApp/rg2-merge-23.05

feat: merge 23.05 to main
NetApp · May 3, 2023 · 34be25d · 34be25d
2 parents b0f9af3 + ec7dd43
commit 34be25d
Show file tree

Hide file tree

Showing 20 changed files with 506 additions and 151 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
diff --git a/cmd/collectors/ems/ems.go b/cmd/collectors/ems/ems.go
@@ -9,6 +9,7 @@ import (
 	"github.com/netapp/harvest/v2/cmd/tools/rest"
 	"github.com/netapp/harvest/v2/pkg/errs"
 	"github.com/netapp/harvest/v2/pkg/matrix"
+	"github.com/netapp/harvest/v2/pkg/set"
 	"github.com/netapp/harvest/v2/pkg/tree/node"
 	"github.com/netapp/harvest/v2/pkg/util"
 	"github.com/tidwall/gjson"
@@ -38,7 +39,7 @@ type Ems struct {
 	DefaultLabels  []string
 	severityFilter string
 	eventNames     []string            // consist of all ems events supported
-	bookendEmsMap  map[string][]string // This is reverse bookend ems map, [Resolving ems]:[Issuing ems slice]
+	bookendEmsMap  map[string]*set.Set // This is reverse bookend ems map, [Resolving ems]:[Set of Issuing ems]. Using Set here to ensure that it has slice of unique issuing ems
 	resolveAfter   time.Duration
 }
 
@@ -93,7 +94,7 @@ func (e *Ems) Init(a *collector.AbstractCollector) error {
 	// init ems props
 	e.InitEmsProp()
 
-	e.bookendEmsMap = make(map[string][]string)
+	e.bookendEmsMap = make(map[string]*set.Set)
 
 	if err = e.InitClient(); err != nil {
 		return err
@@ -323,7 +324,7 @@ func (e *Ems) PollInstance() (map[string]*matrix.Matrix, error) {
 
 	// check instance timestamp and remove it after given resolve_after duration and warning when total instance in cache > 1000 instance
 	for _, issuingEmsList := range e.bookendEmsMap {
-		for _, issuingEms := range issuingEmsList {
+		for _, issuingEms := range issuingEmsList.Slice() {
 			if mx := e.Matrix[issuingEms]; mx != nil {
 				for instanceKey, instance := range mx.GetInstances() {
 					if metr, ok = mx.GetMetrics()["timestamp"]; !ok {
@@ -531,7 +532,7 @@ func (e *Ems) HandleResults(result []gjson.Result, prop map[string][]*emsProp) (
 			     - loop would iterate for all possible issuing ems
 			     - if one or more issuing ems exist then resolve all matching ems else log warning as unable to find matching ems in cache
 			*/
-			for _, issuingEms := range issuingEmsList {
+			for _, issuingEms := range issuingEmsList.Slice() {
 				if mx = m[issuingEms]; mx != nil {
 					metr, exist := mx.GetMetrics()["events"]
 					if !exist {
@@ -557,7 +558,7 @@ func (e *Ems) HandleResults(result []gjson.Result, prop map[string][]*emsProp) (
 			}
 
 			if !emsResolved {
-				e.Logger.Warn().Str("resolving ems", msgName).Str("issue ems", strings.Join(issuingEmsList, ",")).
+				e.Logger.Warn().Str("resolving ems", msgName).Str("issue ems", strings.Join(issuingEmsList.Slice(), ",")).
 					Msg("Unable to find matching issue ems in cache")
 			}
 		} else {
@@ -700,7 +701,7 @@ func (e *Ems) updateMatrix() {
 	tempMap := make(map[string]*matrix.Matrix)
 	// store the bookend ems metric in tempMap
 	for _, issuingEmsList := range e.bookendEmsMap {
-		for _, issuingEms := range issuingEmsList {
+		for _, issuingEms := range issuingEmsList.Slice() {
 			if mx, exist := e.Matrix[issuingEms]; exist {
 				tempMap[issuingEms] = mx
 			}
@@ -715,6 +716,7 @@ func (e *Ems) updateMatrix() {
 	for issuingEms, mx := range tempMap {
 		if metr, ok = mx.GetMetrics()["events"]; !ok {
 			e.Logger.Error().
+				Str("issuingEms", issuingEms).
 				Str("name", "events").
 				Msg("failed to get metric")
 			continue

diff --git a/cmd/collectors/ems/templating.go b/cmd/collectors/ems/templating.go
@@ -1,6 +1,7 @@
 package ems
 
 import (
+	"github.com/netapp/harvest/v2/pkg/set"
 	"github.com/netapp/harvest/v2/pkg/tree/node"
 	"github.com/netapp/harvest/v2/pkg/util"
 	"time"
@@ -151,6 +152,10 @@ func (e *Ems) ParseResolveEms(resolveEvent *node.Node, issueEmsProp emsProp) {
 	}
 	e.Logger.Debug().Str("bookend ems resolve After", e.resolveAfter.String()).Msg("")
 
-	e.bookendEmsMap[resolveEmsName] = append(e.bookendEmsMap[resolveEmsName], issueEmsProp.Name)
+	// Using Set to ensure it has slice of unique issuing ems
+	if _, ok := e.bookendEmsMap[resolveEmsName]; !ok {
+		e.bookendEmsMap[resolveEmsName] = set.New()
+	}
+	e.bookendEmsMap[resolveEmsName].Add(issueEmsProp.Name)
 	e.emsProp[resolveEmsName] = append(e.emsProp[resolveEmsName], &prop)
 }
diff --git a/cmd/collectors/rest/plugins/qtree/qtree.go b/cmd/collectors/rest/plugins/qtree/qtree.go
@@ -230,13 +230,14 @@ func (my Qtree) handlingHistoricalMetrics(result []gjson.Result, data *matrix.Ma
 			my.Logger.Debug().Msgf("add (%s) quota instance: %s.%s.%s.%s", quotaInstanceKey, vserver, volume, tree, qIndex)
 		}
 
-		qtreeInstance := data.GetInstance(vserver + volume + tree)
+		// qtree instancekey would be qtree, svm and volume(sorted keys)
+		qtreeInstance := data.GetInstance(tree + vserver + volume)
 		if qtreeInstance == nil {
 			my.Logger.Warn().
 				Str("tree", tree).
 				Str("volume", volume).
 				Str("vserver", vserver).
-				Msg("No instance matching tree.volume.vserver")
+				Msg("No instance matching tree.vserver.volume")
 			continue
 		}
 

diff --git a/cmd/collectors/zapi/plugins/qtree/qtree.go b/cmd/collectors/zapi/plugins/qtree/qtree.go
@@ -257,7 +257,7 @@ func (my Qtree) handlingHistoricalMetrics(quotas []*node.Node, data *matrix.Matr
 				if my.client.IsClustered() {
 					qtreeInstance = data.GetInstance(tree + "." + volume + "." + vserver)
 				} else {
-					qtreeInstance = data.GetInstance(volume + "." + tree)
+					qtreeInstance = data.GetInstance(tree + "." + volume)
 				}
 				if qtreeInstance == nil {
 					my.Logger.Warn().

diff --git a/cmd/tools/generate/counter.go b/cmd/tools/generate/counter.go
@@ -35,6 +35,14 @@ var (
 		"workload_detail.yaml":        {},
 		"workload_detail_volume.yaml": {},
 	}
+	excludeCounters = map[string]struct{}{
+		"write_latency_histogram": {},
+		"read_latency_histogram":  {},
+		"latency_histogram":       {},
+		"nfsv3_latency_hist":      {},
+		"read_latency_hist":       {},
+		"write_latency_hist":      {},
+	}
 )
 
 type Counters struct {
@@ -243,6 +251,9 @@ func processRestConfigCounters(path string) map[string]Counter {
 	for _, c := range templateCounters.GetAllChildContentS() {
 		if c != "" {
 			name, display, m, _ := util.ParseMetric(c)
+			if _, ok := excludeCounters[name]; ok {
+				continue
+			}
 			description := searchDescriptionSwagger(object, name)
 			harvestName := strings.Join([]string{object, display}, "_")
 			if m == "float" {
@@ -341,6 +352,9 @@ func processZAPIPerfCounters(path string, client *zapi.Client) map[string]Counte
 			}
 			harvestName := strings.Join([]string{object, display}, "_")
 			if m == "float" {
+				if _, ok := excludeCounters[name]; ok {
+					continue
+				}
 				if zapiTypeMap[name] != "string" {
 					co := Counter{
 						Name:        harvestName,
@@ -393,6 +407,9 @@ func processZapiConfigCounters(path string) map[string]Counter {
 	}
 
 	for k, v := range zc {
+		if _, ok := excludeCounters[k]; ok {
+			continue
+		}
 		co := Counter{
 			Name: k,
 			APIs: []MetricDef{
@@ -596,6 +613,10 @@ func processRestPerfCounters(path string, client *rest.Client) map[string]Counte
 			return true
 		}
 		ontapCounterName := r.Get("name").String()
+		if _, ok := excludeCounters[ontapCounterName]; ok {
+			return true
+		}
+
 		description := r.Get("description").String()
 		ty := r.Get("type").String()
 		if override != nil {

diff --git a/cmd/tools/generate/counter.yaml b/cmd/tools/generate/counter.yaml
@@ -248,10 +248,26 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
+
+  - Name: metadata_collector_calc_time
+    Description: amount of time it took to compute metrics between two successive polls, specifically using properties like raw, delta, rate, average, and percent. This metric is available for ZapiPerf/RestPerf collectors.
+    APIs:
+      - API: REST
+        Endpoint: NA
+        ONTAPCounter: Harvest generated
+        Template: NA
+        Unit: microseconds
+      - API: ZAPI
+        Endpoint: NA
+        ONTAPCounter: Harvest generated
+        Template: NA
+        Unit: microseconds
 
   - Name: metadata_collector_instances
     Description: number of objects collected from monitored cluster
@@ -260,10 +276,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
 
   - Name: metadata_collector_metrics
     Description: number of counters collected from monitored cluster
@@ -272,10 +290,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
 
   - Name: metadata_collector_parse_time
     Description: amount of time to parse XML, JSON, etc. for cluster object
@@ -284,10 +304,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
 
   - Name: metadata_collector_plugin_time
     Description: amount of time for all plugins to post-process metrics
@@ -296,10 +318,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
 
   - Name: metadata_collector_poll_time
     Description: amount of time it took for the poll to finish
@@ -308,10 +332,26 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
+      - API: ZAPI
+        Endpoint: NA
+        ONTAPCounter: Harvest generated
+        Template: NA
+        Unit: microseconds
+
+  - Name: metadata_collector_skips
+    Description: number of metrics that were not calculated between two successive polls. This metric is available for ZapiPerf/RestPerf collectors.
+    APIs:
+      - API: REST
+        Endpoint: NA
+        ONTAPCounter: Harvest generated
+        Template: NA
+        Unit: scalar
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
 
   - Name: metadata_collector_task_time
     Description: amount of time it took for each collector's subtasks to complete
@@ -320,10 +360,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
 
   - Name: metadata_component_count
     Description: number of metrics collected for each object
@@ -332,10 +374,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
 
   - Name: metadata_component_status
     Description: status of the collector - 0 means running, 1 means standby, 2 means
@@ -345,10 +389,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: enum
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: enum
 
   - Name: metadata_exporter_count
     Description: number of metrics and labels exported
@@ -357,10 +403,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
 
   - Name: metadata_exporter_time
     Description: amount of time it took to render, export, and serve exported data
@@ -369,10 +417,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: microseconds
 
   - Name: metadata_target_goroutines
     Description: number of goroutines that exist within the poller
@@ -381,10 +431,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: scalar
 
   - Name: metadata_target_status
     Description: status of the system being monitored. 0 means reachable, 1 means unreachable
@@ -393,10 +445,12 @@ counters:
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: enum
       - API: ZAPI
         Endpoint: NA
         ONTAPCounter: Harvest generated
         Template: NA
+        Unit: enum
 
   - Name: node_cpu_busytime
     Description: The time (in hundredths of a second) that the CPU has been doing useful
@@ -414,10 +468,12 @@ counters:
         Endpoint: api/cluster/counter/tables/qos_detail
         ONTAPCounter: Harvest generated
         Template: conf/restperf/9.12.0/workload_detail.yaml
+        Unit: microseconds
       - API: ZAPI
         Endpoint: perf-object-get-instances workload_detail
         ONTAPCounter: Harvest generated
         Template: conf/zapiperf/9.12.0/workload_detail.yaml
+        Unit: microseconds
 
   - Name: qos_detail_volume_resource_latency
     Description: average latency for volume on Data ONTAP subsystems
@@ -426,10 +482,12 @@ counters:
         Endpoint: api/cluster/counter/tables/qos_detail_volume
         ONTAPCounter: Harvest generated
         Template: conf/restperf/9.12.0/workload_detail_volume.yaml
+        Unit: microseconds
       - API: ZAPI
         Endpoint: perf-object-get-instances workload_detail_volume
         ONTAPCounter: Harvest generated
         Template: conf/zapiperf/9.12.0/workload_detail_volume.yaml
+        Unit: microseconds
 
   - Name: quota_disk_limit
     Description: Maximum amount of disk space, in kilobytes, allowed for the quota target