Merge branch 'master' into clean-up-script-names

README.md

@@ -2,4 +2,18 @@
 
 > A theodolite is a precision optical instrument for measuring angles between designated visible points in the horizontal and vertical planes.  -- <cite>[Wikipedia](https://en.wikipedia.org/wiki/Theodolite)</cite>
 
-Theodolite is a framework for benchmarking the horizontal and vertical scalability of stream processing engines.
+Theodolite is a framework for benchmarking the horizontal and vertical scalability of stream processing engines. It consists of three modules:
+
+## Theodolite Benchmarks
+
+Theodolite contains 4 application benchmarks, which are based on typical use cases for stream processing within microservices. For each benchmark, a corresponding workload generator is provided. Currently, this repository provides benchmark implementations for Kafka Streams.
+
+
+## Theodolite Execution Framework
+
+Theodolite aims to benchmark scalability of stream processing engines for real use cases. Microservices that apply stream processing techniques are usually deployed in elastic cloud environments. Hence, Theodolite's cloud-native benchmarking framework deploys as components in a cloud environment, orchestrated by Kubernetes. More information on how to execute scalability benchmarks can be found in [Thedolite execution framework](execution).
+
+
+## Theodolite Analysis Tools
+
+Theodolite's benchmarking method create a *scalability graph* allowing to draw conclusions about the scalability of a stream processing engine or its deployment. A scalability graph shows how resource demand evolves with an increasing workload. Theodolite provides Jupyter notebooks for creating such scalability graphs based on benchmarking results from the execution framework. More information can be found in [Theodolite analysis tool](analysis).

analysis/README.md

@@ -17,6 +17,6 @@ Python libraries, which can be installed via:
 pip install -r requirements.txt 
 ```
 
- We have tested these
+We have tested these
 notebooks with [Visual Studio Code](https://code.visualstudio.com/docs/python/jupyter-support),
 however, every other server should be fine as well.

execution/.gitignore

+exp_counter.txt
\ No newline at end of file

execution/README.md

@@ -95,7 +95,15 @@ kubectl apply -f infrastructure/kafka/service-monitor.yaml
 
 ##### Other options for Kafka
 
-Other Kafka deployments, for example, using Strimzi, should work in similiar way.
+Other Kafka deployments, for example, using Strimzi, should work in a similar way.
+
+#### A Kafka Client Pod
+
+A permanently running pod used for Kafka configuration is started via:
+
+```sh
+kubectl apply -f infrastructure/kafka/kafka-client.yaml 
+```
 
 #### The Kafka Lag Exporter
 

execution/execution.sh

-#!/bin/bash
-
-./run_loop.sh 1 "25000 50000 75000 100000 125000 150000" "1 2 3 4 5" 40 #6*5=3Std
-sleep 5m
-./run_loop.sh 2 "6 7 8 9" "1 2 3 4 6 8 10 12 14 16 18 20" 40 #4*12=5Std
-sleep 5m
-./run_loop.sh 3 "25000 50000 75000 100000 125000 150000" "1 2 3 4 5 6" 40 #6*6=3.5Std
-sleep 5m
-./run_loop.sh 4 "25000 50000 75000 100000 125000 150000" "1 2 4 6 8 10 12 14 16 18 20 30 40 50 60 70 80 90" 40 #6*18=11Std
-sleep 5m
-
-./run_loop.sh 1 "25000 50000 75000 100000 125000 150000" "1 2 3 4 5" 400 #6*5=3Std
-sleep 5m
-./run_loop.sh 2 "6 7 8 9" "1 2 3 4 6 8 10 12 14 16 18 20" 400 #4*12=5Std
-sleep 5m
-./run_loop.sh 3 "25000 50000 75000 100000 125000 150000" "1 2 3 4 5 6" 400 #6*6=3.5Std
-sleep 5m
-./run_loop.sh 4 "25000 50000 75000 100000 125000 150000" "1 2 4 6 8 10 12 14 16 18 20 30 40 50 60 70 80 90" 400 #6*18=11Std
-sleep 5m
-./run_loop.sh 4 "150000" "100 110 120 130 140 150 160 17 18 190 200" 400 #6*18=11Std
-sleep 5m
-# For commit interval evaluation
-./run_loop.sh 4 "5000 10000 15000 20000 25000 30000" "1 2 3 4 5 6 7 8 9 10 11 12 13 14 15" 160
\ No newline at end of file

execution/execution_tmp_200507.sh

-#!/bin/bash
-
-#./run_loop.sh 1 "50000 100000 150000 200000 250000 300000" "1 2 3 4 5" 40 #3Std
-./run_loop.sh 1 "200000 250000 300000" "1 2 3 4 5" 40 1000m 4Gi 100 5 #1.5Std
-sleep 1m
-#./run_loop.sh 1 "50000 100000 150000 200000 250000 300000" "1 2 3 4 5" 400 #3Std
-./run_loop.sh 1 "200000 250000 300000" "1 2 3 4 5" 400 1000m 4Gi 100 5 #1.5Std
-sleep 1m
-
-#./run_loop.sh 3 "50000 100000 150000 200000 250000 300000" "1 2 3 4 5 6 7 8 9 10" 40 #6 Std
-./run_loop.sh 3 "200000 250000 300000" "1 2 3 4 5 6 7 8 9 10" 40 1000m 4Gi 100 5 #3 Std
-sleep 1m
-#./run_loop.sh 3 "50000 100000 150000 200000 250000 300000" "1 2 3 4 5 6 7 8 9 10" 400 #6 Std
-./run_loop.sh 3 "200000 250000 300000" "1 2 3 4 5 6 7 8 9 10" 400 1000m 4Gi 100 5 #3 Std
-sleep 1m
-
-./run_loop.sh 1 "50000 100000 150000 200000 250000 300000" "1 2 3 4 5" 40 500m 2Gi 100 5 #3Std

execution/exp_counter.txt

-0
+1

execution/infrastructure/kafka/kafka-client.yaml

+apiVersion: v1
+kind: Pod
+metadata:
+  name: kafka-client-2
+spec:
+  containers:
+  - name: kafka-client
+    image: confluentinc/cp-enterprise-kafka:5.4.0
+    command:
+      - sh
+      - -c
+      - "exec tail -f /dev/null"
\ No newline at end of file

execution/run_uc1.sh

@@ -34,7 +34,11 @@ echo "$WORKLOAD_GENERATOR_YAML" | kubectl apply -f -
 
 # Start application
 REPLICAS=$INSTANCES
-#kubectl apply -f uc3-application/aggregation-deployment.yaml
+# When not using `sed` anymore, use `kubectl apply -f uc1-application`
+kubectl apply -f uc1-application/aggregation-service.yaml
+kubectl apply -f uc1-application/jmx-configmap.yaml
+kubectl apply -f uc1-application/service-monitor.yaml
+#kubectl apply -f uc1-application/aggregation-deployment.yaml
 APPLICATION_YAML=$(sed "s/{{CPU_LIMIT}}/$CPU_LIMIT/g; s/{{MEMORY_LIMIT}}/$MEMORY_LIMIT/g; s/{{KAFKA_STREAMS_COMMIT_INTERVAL_MS}}/$KAFKA_STREAMS_COMMIT_INTERVAL_MS/g" uc1-application/aggregation-deployment.yaml)
 echo "$APPLICATION_YAML" | kubectl apply -f -
 kubectl scale deployment titan-ccp-aggregation --replicas=$REPLICAS
@@ -52,6 +56,9 @@ deactivate
 #sed "s/{{INSTANCES}}/1/g" uc1-workload-generator/deployment.yaml | kubectl delete -f -
 #sed "s/{{NUM_SENSORS}}/$NUM_SENSORS/g; s/{{INSTANCES}}/$WL_INSTANCES/g" uc1-workload-generator/deployment.yaml | kubectl delete -f -
 echo "$WORKLOAD_GENERATOR_YAML" | kubectl delete -f -
+kubectl delete -f uc1-application/aggregation-service.yaml
+kubectl delete -f uc1-application/jmx-configmap.yaml
+kubectl delete -f uc1-application/service-monitor.yaml
 #kubectl delete -f uc1-application/aggregation-deployment.yaml
 echo "$APPLICATION_YAML" | kubectl delete -f -
 

execution/run_uc2.sh

@@ -30,6 +30,10 @@ sed "s/{{NUM_NESTED_GROUPS}}/$NUM_NESTED_GROUPS/g" uc2-workload-generator/deploy
 
 # Start application
 REPLICAS=$INSTANCES
+# When not using `sed` anymore, use `kubectl apply -f uc2-application`
+kubectl apply -f uc2-application/aggregation-service.yaml
+kubectl apply -f uc2-application/jmx-configmap.yaml
+kubectl apply -f uc2-application/service-monitor.yaml
 #kubectl apply -f uc2-application/aggregation-deployment.yaml
 APPLICATION_YAML=$(sed "s/{{CPU_LIMIT}}/$CPU_LIMIT/g; s/{{MEMORY_LIMIT}}/$MEMORY_LIMIT/g; s/{{KAFKA_STREAMS_COMMIT_INTERVAL_MS}}/$KAFKA_STREAMS_COMMIT_INTERVAL_MS/g" uc2-application/aggregation-deployment.yaml)
 echo "$APPLICATION_YAML" | kubectl apply -f -
@@ -45,6 +49,9 @@ deactivate
 
 # Stop wl and app
 kubectl delete -f uc2-workload-generator/deployment.yaml
+kubectl delete -f uc2-application/aggregation-service.yaml
+kubectl delete -f uc2-application/jmx-configmap.yaml
+kubectl delete -f uc2-application/service-monitor.yaml
 #kubectl delete -f uc2-application/aggregation-deployment.yaml
 echo "$APPLICATION_YAML" | kubectl delete -f -
 

execution/run_uc3.sh

@@ -34,6 +34,10 @@ echo "$WORKLOAD_GENERATOR_YAML" | kubectl apply -f -
 
 # Start application
 REPLICAS=$INSTANCES
+# When not using `sed` anymore, use `kubectl apply -f uc3-application`
+kubectl apply -f uc3-application/aggregation-service.yaml
+kubectl apply -f uc3-application/jmx-configmap.yaml
+kubectl apply -f uc3-application/service-monitor.yaml
 #kubectl apply -f uc3-application/aggregation-deployment.yaml
 APPLICATION_YAML=$(sed "s/{{CPU_LIMIT}}/$CPU_LIMIT/g; s/{{MEMORY_LIMIT}}/$MEMORY_LIMIT/g; s/{{KAFKA_STREAMS_COMMIT_INTERVAL_MS}}/$KAFKA_STREAMS_COMMIT_INTERVAL_MS/g" uc3-application/aggregation-deployment.yaml)
 echo "$APPLICATION_YAML" | kubectl apply -f -
@@ -51,7 +55,10 @@ deactivate
 #kubectl delete -f uc3-workload-generator/deployment.yaml
 #sed "s/{{INSTANCES}}/1/g" uc3-workload-generator/deployment.yaml | kubectl delete -f -
 echo "$WORKLOAD_GENERATOR_YAML" | kubectl delete -f -
-#kubectl delete -f uc1-application/aggregation-deployment.yaml
+kubectl delete -f uc3-application/aggregation-service.yaml
+kubectl delete -f uc3-application/jmx-configmap.yaml
+kubectl delete -f uc3-application/service-monitor.yaml
+#kubectl delete -f uc3-application/aggregation-deployment.yaml
 #sed "s/{{CPU_LIMIT}}/1000m/g; s/{{MEMORY_LIMIT}}/4Gi/g; s/{{KAFKA_STREAMS_COMMIT_INTERVAL_MS}}/100/g" uc3-application/aggregation-deployment.yaml | kubectl delete -f -
 echo "$APPLICATION_YAML" | kubectl delete -f -
 

execution/run_uc4.sh

@@ -32,6 +32,10 @@ sed "s/{{NUM_SENSORS}}/$NUM_SENSORS/g" uc4-workload-generator/deployment.yaml |
 # Start application
 REPLICAS=$INSTANCES
 #AGGREGATION_DURATION_DAYS=$DIM_VALUE
+# When not using `sed` anymore, use `kubectl apply -f uc4-application`
+kubectl apply -f uc4-application/aggregation-service.yaml
+kubectl apply -f uc4-application/jmx-configmap.yaml
+kubectl apply -f uc4-application/service-monitor.yaml
 #kubectl apply -f uc4-application/aggregation-deployment.yaml
 #sed "s/{{AGGREGATION_DURATION_DAYS}}/$AGGREGATION_DURATION_DAYS/g" uc4-application/aggregation-deployment.yaml | kubectl apply -f -
 APPLICATION_YAML=$(sed "s/{{CPU_LIMIT}}/$CPU_LIMIT/g; s/{{MEMORY_LIMIT}}/$MEMORY_LIMIT/g; s/{{KAFKA_STREAMS_COMMIT_INTERVAL_MS}}/$KAFKA_STREAMS_COMMIT_INTERVAL_MS/g" uc4-application/aggregation-deployment.yaml)
@@ -48,6 +52,9 @@ deactivate
 
 # Stop wl and app
 kubectl delete -f uc4-workload-generator/deployment.yaml
+kubectl delete -f uc4-application/aggregation-service.yaml
+kubectl delete -f uc4-application/jmx-configmap.yaml
+kubectl delete -f uc4-application/service-monitor.yaml
 #kubectl delete -f uc4-application/aggregation-deployment.yaml
 echo "$APPLICATION_YAML" | kubectl delete -f -
 

execution/theodolite.sh

@@ -10,7 +10,7 @@ KAFKA_STREAMS_COMMIT_INTERVAL_MS=${7:-100}
 EXECUTION_MINUTES=${8:-5}
 
 # Get and increment counter
-EXP_ID=$(cat exp_counter.txt)
+EXP_ID=$(cat exp_counter.txt 2>/dev/null || echo "0")
 echo $((EXP_ID+1)) > exp_counter.txt
 
 # Store meta information

execution/uc1-workload-generator/jmx-configmap.yaml

-apiVersion: v1
-kind: ConfigMap
-metadata:
-  name: load-generator-jmx-configmap
-data:
-  jmx-kafka-prometheus.yml: |+
-    jmxUrl: service:jmx:rmi:///jndi/rmi://localhost:5555/jmxrmi
-    lowercaseOutputName: true
-    lowercaseOutputLabelNames: true
-    ssl: false

execution/uc2-workload-generator/jmx-configmap.yaml

-apiVersion: v1
-kind: ConfigMap
-metadata:
-  name: load-generator-jmx-configmap
-data:
-  jmx-kafka-prometheus.yml: |+
-    jmxUrl: service:jmx:rmi:///jndi/rmi://localhost:5555/jmxrmi
-    lowercaseOutputName: true
-    lowercaseOutputLabelNames: true
-    ssl: false

execution/uc3-workload-generator/jmx-configmap.yaml

-apiVersion: v1
-kind: ConfigMap
-metadata:
-  name: load-generator-jmx-configmap
-data:
-  jmx-kafka-prometheus.yml: |+
-    jmxUrl: service:jmx:rmi:///jndi/rmi://localhost:5555/jmxrmi
-    lowercaseOutputName: true
-    lowercaseOutputLabelNames: true
-    ssl: false

execution/uc4-workload-generator/jmx-configmap.yaml

-apiVersion: v1
-kind: ConfigMap
-metadata:
-  name: load-generator-jmx-configmap
-data:
-  jmx-kafka-prometheus.yml: |+
-    jmxUrl: service:jmx:rmi:///jndi/rmi://localhost:5555/jmxrmi
-    lowercaseOutputName: true
-    lowercaseOutputLabelNames: true
-    ssl: false