Dogfood e2etest; Client/Server Application

dogfood/CONTRIBUTING.md

@@ -25,3 +25,42 @@ If your changes don't seem to propagate, you can:
 - `make uninstall` and `make install`
   or move to the top level directory and run
 - `colima delete` and `make colima-start` and redo [dogfood instructions](README.md)
+
+## Testing Datadog Metrics
+
+To apply the datadog agent to your local colima environment, run the following:
+
+```
+kubectl apply -f "https://raw.githubusercontent.com/DataDog/datadog-agent/master/Dockerfiles/manifests/rbac/clusterrole.yaml"
+
+kubectl apply -f "https://raw.githubusercontent.com/DataDog/datadog-agent/master/Dockerfiles/manifests/rbac/serviceaccount.yaml"
+
+kubectl apply -f "https://raw.githubusercontent.com/DataDog/datadog-agent/master/Dockerfiles/manifests/rbac/clusterrolebinding.yaml"
+```
+
+Then take a look at the file `datadog-agent-all-features.yaml` (Feel free to remove the SECURITY feature as it is 
+unnecessary for testing). You will notice that an api key AND a random string encoded in base64 is required. Get yourself
+an API key from your Datadog site, think of a random string, then do the following:
+
+```
+echo -n '<Your API key>' | base64
+# Copy the encoding and paste it where needed in the datadog.yaml
+echo -n 'Random string' | base64
+# Copy the encoding and paste it where needed in the datadog.yaml
+```
+
+By default the Datadog site is set to the US site datadoghq.com. If you're using other sites, you may want to edit the
+`DD_SITE` environment variable accordingly.
+
+Deploy the Daemonset:
+```
+kubectl apply -f datadog-agent-all-features.yaml
+```
+
+Verify it is running correctly using `kubectl get daemonset` in the appropriate namespace (`default` is the default)
+
+Once you've verified the daemonset is up and running, you'll need to get Kubernetes State Metrics with the following steps:
+1. Download the kube-state manifests folder [here](https://github.com/kubernetes/kube-state-metrics/tree/master/examples/standard).
+2. `kubectl apply -f <NAME_OF_THE_KUBE_STATE_MANIFESTS_FOLDER>`
+
+Then you should be set to see metrics for the client and server containers.

dogfood/Makefile

@@ -41,7 +41,8 @@ install:
 uninstall:
 	helm template ./client/chart | kubectl delete -f -
 	helm template ./server/chart | kubectl delete -f -
-	kubectl delete -f ../examples/namespace.yaml
+
+reinstall: uninstall install
 
 restart-client:
 	kubectl -n chaos-demo rollout restart deployment chaos-dogfood-client

dogfood/README.md

@@ -1,4 +1,4 @@
-# Installing protoc
+ # Installing protoc
 
 Run `brew install protobuf` or `make install-protobuf`
 
@@ -104,6 +104,13 @@ x
 You can `kubectl apply -f examples/<disruption.yaml>` for any `example/` disruption files.
 For gRPC disruption, you can follow these [detailed steps](../docs/grpc_disruption/demo_instructions.md).
 
+### Sending Metrics to Datadog
+
+For the purposes of testing disruptions/workflows, you should make sure that the datadog agent is properly installed
+on the cluster that the client and server are running on. 3 of the major disruptive resources properly send metrics
+to Datadog (CPU, Network, Disk). The client contains computation related to these disruptions and can be tested using
+the disruptions mentioned.
+
 ### Clean up
 
 - Run `make uninstall` to `kubectl delete` both charts as well as remove the namespace.

dogfood/client/chart/templates/deployment.yaml

@@ -24,7 +24,22 @@ spec:
       {{- end }}
       securityContext:
         {{- toYaml .Values.podSecurityContext | nindent 8 }}
+      volumes:
+        - name: data
+          persistentVolumeClaim:
+            claimName: dogfood-client-pvc
       containers:
+        - name: read-file
+          image: ubuntu:bionic-20220128
+          command: [ "/bin/bash" ]
+          args:
+            [
+              "-c",
+              "echo 'create file to read from: /mnt/data/disk-read-file' && dd if=/dev/zero of=/mnt/data/disk-read-file bs=20k count=1; while true; do time dd if=/mnt/data/disk-read-file of=/dev/null iflag=direct; sleep 1; done",
+            ]
+          volumeMounts:
+              - mountPath: /mnt/data
+                name: data
         - name: {{ .Chart.Name }}
           securityContext:
             {{- toYaml .Values.securityContext | nindent 12 }}
@@ -39,3 +54,6 @@ spec:
               protocol: TCP
           resources:
             {{- toYaml .Values.resources | nindent 12 }}
+          volumeMounts:
+            - mountPath: /mnt/data
+              name: data

dogfood/client/chart/templates/volumeclaim.yaml

@@ -0,0 +1,17 @@
+# Unless explicitly stated otherwise all files in this repository are licensed
+# under the Apache License Version 2.0.
+# This product includes software developed at Datadog (https://www.datadoghq.com/).
+# Copyright 2022 Datadog, Inc.
+
+apiVersion: v1
+kind: PersistentVolumeClaim
+metadata:
+  name: dogfood-client-pvc
+  namespace: chaos-demo
+spec:
+  storageClassName: longhorn
+  accessModes:
+    - ReadWriteOnce
+  resources:
+    requests:
+      storage: 3Gi

dogfood/client/dogfood_client.go

@@ -7,9 +7,11 @@ package main
 
 import (
 	"context"
+	"errors"
 	"flag"
 	"fmt"
 	"log"
+	"os"
 	"strconv"
 	"time"
 
@@ -56,7 +58,50 @@ func getCatalogWithTimeout(client pb.ChaosDogfoodClient) ([]*pb.CatalogItem, err
 	return res.Items, nil
 }
 
-// regularly order food for different aniamls
+func printAndLog(logLine string) {
+	fmt.Println(logLine)
+
+	go func() {
+		f, err := os.OpenFile("/dev/urandom", os.O_RDONLY|os.O_SYNC, 0644)
+		if err != nil {
+			log.Fatal(err)
+		}
+
+		logLineBytes := make([]byte, 500000)
+
+		_, err = f.Read(logLineBytes)
+		if err != nil {
+			log.Fatal(err)
+		}
+
+		if err = f.Close(); err != nil {
+			log.Fatal(err)
+		}
+
+		if _, err = os.Stat("/mnt/data/logging"); errors.Is(err, os.ErrNotExist) {
+			f, err = os.Create("/mnt/data/logging")
+			if err != nil {
+				log.Fatal(err)
+			}
+		} else {
+			f, err = os.OpenFile("/mnt/data/logging", os.O_WRONLY|os.O_SYNC, 0644)
+			if err != nil {
+				log.Fatal(err)
+			}
+		}
+
+		_, err = f.Write(logLineBytes)
+		if err != nil {
+			log.Fatal(err)
+		}
+
+		if err = f.Close(); err != nil {
+			log.Fatal(err)
+		}
+	}()
+}
+
+// regularly order food for different animals
 // note: mouse should return error because food for mice is not in the catalog
 func sendsLotsOfRequests(client pb.ChaosDogfoodClient) {
 	animals := []string{"dog", "cat", "mouse"}
@@ -66,24 +111,24 @@ func sendsLotsOfRequests(client pb.ChaosDogfoodClient) {
 
 	for {
 		// visually mark a new loop in logs
-		fmt.Println("x")
+		printAndLog("x")
 
 		// grab catalog
 		items, err := getCatalogWithTimeout(client)
 		if err != nil {
-			fmt.Printf("| ERROR getting catalog:%v\n", err.Error())
+			printAndLog(fmt.Sprintf("| ERROR getting catalog:%v\n", err.Error()))
 		}
 
-		fmt.Printf("| catalog: %v items returned %s\n", strconv.Itoa(len(items)), stringifyCatalogItems(items))
+		printAndLog(fmt.Sprintf("| catalog: %v items returned %s\n", strconv.Itoa(len(items)), stringifyCatalogItems(items)))
 		time.Sleep(time.Second)
 
 		// make an order
 		order, err := orderWithTimeout(client, animals[i])
 		if err != nil {
-			fmt.Printf("| ERROR ordering food: %v\n", err.Error())
+			printAndLog(fmt.Sprintf("| ERROR ordering food: %v\n", err.Error()))
 		}
 
-		fmt.Printf("| ordered: %v\n", order)
+		printAndLog(fmt.Sprintf("| ordered: %v\n", order))
 		time.Sleep(time.Second)
 
 		// iterate
@@ -106,9 +151,10 @@ func stringifyCatalogItems(items []*pb.CatalogItem) string {
 
 func main() {
 	// create and eventually close connection
-	fmt.Printf("connecting to %v...\n", serverAddr)
+	printAndLog(fmt.Sprintf("connecting to %v...\n", serverAddr))
 
 	var opts []grpc.DialOption
+
 	opts = append(opts, grpc.WithInsecure())
 	opts = append(opts, grpc.WithBlock())
 
@@ -126,5 +172,7 @@ func main() {
 	// generate and use client
 	client := pb.NewChaosDogfoodClient(conn)
 
+	printAndLog("We successfully generated the client, getting ready to send requests")
+
 	sendsLotsOfRequests(client)
 }