Modified how logging with monitord container is implemented, with simpler logic thanks to the argo client library and k8 client-go for pods' logs

logger/argo_logs.go

@@ -0,0 +1,219 @@
+package logger
+
+import (
+	"bufio"
+	"encoding/json"
+	"fmt"
+	"oc-monitord/tools"
+	"oc-monitord/utils"
+	"slices"
+	"time"
+
+	"github.com/rs/zerolog"
+	"k8s.io/apimachinery/pkg/watch"
+
+	wfv1 "github.com/argoproj/argo-workflows/v3/pkg/apis/workflow/v1alpha1"
+)
+
+// An object to monitor the logs generated by a specific pod from a workflow execution
+type ArgoWatch struct {
+	Name      string
+	Namespace string
+	Status    string
+	Conditions
+	Created  string
+	Started  string
+	Duration string
+	Progress string
+	Logs     []string
+}
+
+type Conditions struct {
+	PodRunning bool
+	Completed  bool
+}
+
+func (a *ArgoWatch) Equals(arg *ArgoWatch) bool {
+	if arg == nil {
+		return false
+	}
+	return a.Status == arg.Status && a.Progress == arg.Progress && a.Conditions.PodRunning == arg.Conditions.PodRunning && a.Conditions.Completed == arg.Conditions.Completed
+}
+
+func NewArgoLogs(name string, namespace string, stepMax int) *ArgoLogs {
+	return &ArgoLogs{
+		Name:        "oc-monitor-" + name,
+		Namespace:   namespace,
+		CreatedDate: time.Now().Format("2006-01-02 15:04:05"),
+		StepCount:   0,
+		StepMax:     stepMax,
+		stop:        false,
+		Seen:        []string{},
+	}
+}
+
+// An object to monitor and log the output of an argo submit 
+type ArgoLogs struct {
+	Name        string
+	Namespace   string
+	CreatedDate string
+	StepCount   int
+	StepMax     int
+	stop        bool
+	Started     time.Time
+	Seen        []string
+	Logs        []string
+	IsStreaming bool
+}
+
+func (a *ArgoLogs) NewWatch() *ArgoWatch {
+	return &ArgoWatch{
+		Name:      a.Name,
+		Namespace: a.Namespace,
+		Status:    "Pending",
+		Created:   a.CreatedDate,
+		Started:   a.Started.Format("2006-01-02 15:04:05"),
+		Conditions: Conditions{
+			PodRunning: a.StepCount > 0 && a.StepCount < a.StepMax,
+			Completed:  a.StepCount == a.StepMax,
+		},
+		Progress: fmt.Sprintf("%v/%v", a.StepCount, a.StepMax),
+		Duration: "0s",
+		Logs:     []string{},
+	}
+
+}
+
+func (a *ArgoLogs) StartStepRecording(current_watch *ArgoWatch, logger zerolog.Logger) {
+	jsonified, _ := json.Marshal(current_watch)
+	logger.Info().Msg(string(jsonified))
+	a.StepCount += 1
+	a.Started = time.Now()
+}
+
+
+type ArgoPodLog struct {
+	PodName		string
+	Step		string
+	Message		string
+}
+
+func NewArgoPodLog(name string, step string, msg string) ArgoPodLog {
+	return ArgoPodLog{
+		PodName: name,
+		Step: step,
+		Message: msg,
+	}
+}
+
+func LogKubernetesArgo(wfName string, executionID string, watcher watch.Interface) {
+	var argoWatcher *ArgoWatch
+	var pods []string
+	var node wfv1.NodeStatus
+
+	wfl := utils.GetWFLogger("")
+
+	for event := range (watcher.ResultChan()) {
+		wf, ok := event.Object.(*wfv1.Workflow)
+		if !ok {
+			wfl.Error().Msg("unexpected type")
+			continue
+		}
+		if len(wf.Status.Nodes) == 0 {
+			wfl.Debug().Msg("No node status yet")	// The first output of the channel doesn't contain Nodes so we skip it
+			continue
+		}
+		
+		conditions := retrieveCondition(wf)	
+
+		// Retrieving the Status for the main node, which is named after the workflow
+		if node, ok = wf.Status.Nodes[wfName]; !ok {
+			bytified, _ := json.MarshalIndent(wf.Status.Nodes,"","\t") 
+			wfl.Fatal().Msg("Could not find the " + wfName + " node in \n" + string(bytified))
+		}
+
+		now := time.Now()
+		start, _ := time.Parse(time.RFC3339, node.StartedAt.String() )
+		duration := now.Sub(start)
+
+		newWatcher := ArgoWatch{
+			Name: node.Name,
+			Namespace: executionID,
+			Status: string(node.Phase),
+			Created: node.StartedAt.String(),
+			Started: node.StartedAt.String(),
+			Progress: string(node.Progress),
+			Duration: duration.String(),
+			Conditions: conditions,
+		}
+
+		if argoWatcher == nil {
+			argoWatcher = &newWatcher
+		}
+
+		if !newWatcher.Equals(argoWatcher){
+			jsonified, _ := json.Marshal(newWatcher)
+			wfl.Info().Msg(string(jsonified))
+			argoWatcher = &newWatcher
+		}
+		
+		// I don't think we need to use WaitGroup here, because the loop itself 
+		// acts as blocking process for the main thread, because Argo watch never closes the channel
+		for _, pod := range wf.Status.Nodes{
+			if !slices.Contains(pods,pod.Name){
+				pl := wfl.With().Str("pod",  pod.Name).Logger()
+				if wfName == pod.Name { pods = append(pods, pod.Name); continue }	// One of the node is the Workflow, the others are the pods so don't try to log on the wf name
+				go logKubernetesPods(executionID, wfName, pod.Name, pl)
+				pods = append(pods, pod.Name)
+			} 
+		}
+
+		// Stop listening to the chan when the Workflow is completed or something bad happened
+		if node.Phase.Completed() {
+			wfl.Info().Msg(wfName + " worflow completed")
+			break
+		}
+		if node.Phase.FailedOrError() {
+			wfl.Error().Msg(wfName + "has failed, please refer to the logs")
+			wfl.Error().Msg(node.Message)	
+			break
+		}
+	}
+}
+
+func retrieveCondition(wf *wfv1.Workflow) (c Conditions) {
+	for _, cond := range wf.Status.Conditions {
+		if cond.Type == "PodRunning" {
+			c.PodRunning = cond.Status == "True"
+		}
+		if cond.Type == "Completed" {
+			c.Completed = cond.Status == "True"
+		}
+	}
+
+	return 
+	
+}
+
+// Function needed to be executed as a go thread 
+func logKubernetesPods(executionId string, wfName string,podName string, logger zerolog.Logger){
+	k, err := tools.NewKubernetesTool()
+	if err != nil {
+		logger.Error().Msg("Could not get Kubernetes tools")
+		return
+	}
+
+	reader, err := k.GetPodLogger(executionId, wfName, podName)
+	if err != nil {
+		logger.Error().Msg(err.Error())
+		return
+	}
+
+	scanner := bufio.NewScanner(reader)
+	for scanner.Scan() {
+		log := scanner.Text()
+		podLog := NewArgoPodLog(wfName,podName,log)
+		jsonified, _ := json.Marshal(podLog)
+		logger.Info().Msg(string(jsonified))
+	}
+}