Summary
Data is king and your users need a modern data platform quickly.
With this Fling, you will leverage your VMware Cloud Foundation 4.0 deployment and stand a sample data platform on a Tanzu Kubernetes Grid guest cluster in less than 20-minutes comprising of Kafka, Spark, Solr, and ELK.
Additionally, this Fling comes with a market data sample application (using real market data from dxFeed) that shows how all these data platform components work together.
The diagram below depicts the final outcome of the Data Platform on VCF4.0:
The sample market data app:
The diagram below depicts the final outcome of the Market Data Sample App:
- Data Platform Provisioning: automation instantiates Kafka, Spark, Solr, ELK in 20 minutes.
- Kafka Publisher: Retrieves delayed market data from publicly available DevExpert dxfeed and publishes it on multiple Kafka topics
- Spark Subscriber : Subscribes to the market data from Kafka topics, manipulates the data and finally writes into Solr.
- Logstash Subscriber : Subscribes to the market data from Kafka topics and writes into Elasticsearch through logstash and also creates a kibana dashboard.
About dxFeed - https://www.dxfeed.com : it is a subsidiary of Devexperts, with the primary focus of delivering financial markets information and services to buy-side and sell-side institutions of the global financial industry, specifically to traders, data analysts, quants and portfolio managers.
Sample Trades data from dxFeed:
{"Trade" : {
"MSFT" : {
"eventSymbol" : "MSFT",
"eventTime" : 0,
"time" : 1598385600578,
"timeNanoPart" : 0,
"sequence" : 200,
"exchangeCode" : "Q",
"price" : 216.47,
"change" : 0.0,
"size" : 2197641.0,
"dayVolume" : 8833.0,
"dayTurnover" : 1917851.3,
"tickDirection" : "ZERO_UP",
"extendedTradingHours" : false
}}} Sample Quotes data from dxFeed:
{"Quote" : { "MSFT" : { "eventSymbol" : "MSFT", "eventTime" : 0, "sequence" : 0, "timeNanoPart" : 0, "bidTime" : 1598433071000, "bidExchangeCode" : "P", "bidPrice" : 217.17, "bidSize" : 3.0, "askTime" : 1598433300000, "askExchangeCode" : "P", "askPrice" : 217.39, "askSize" : 1.0 }}}
Final Output: Kibana Dashboard
Stay tuned for future VMware Tanzu data platforms initiatives.
Part One:
Part Two:
Requirements
- Kubernetes cluster with 1 control plane and 8 worker nodes.
- Control Plane: 1 Small VM (2 CPUs, 2GB RAM)
- Worker Nodes: 8 Large VMs (4 CPUs, 16GB RAM)
Instructions
Sample Data Platform Fling
Data is king and your users need a sample data platform quickly.
With this Fling, you will leverage your VMware Cloud Foundation 4.0 deployment and stand a sample data platform on a Tanzu Kubernetes Grid guest cluster in less than 20-minutes comprising of Kafka, Spark, Solr, and ELK.
Setup Instructions
Run the below commands to untar dataplatforms-fling-package-kubernetes.tar.gz
tar -xvf dataplatforms-fling-package-kubernetes.tar.gz
cd kubernetes
BASE_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" > /dev/null 2>&1 && cd .. && pwd)"
We have divided the Data Platform automation in three parts:
- Infrastructure: Setup of Tanzu Kubernetes Guest Cluster (VMware Cloud Foundatation configured as per VMware Validated Design (VVD)). The final outcome is a k8s cluster with 8 workers as per yaml file.
- Data Platform Components: Installation of Kafka, Spark, Solr and ElasticSearch.
- Market Data Sample App: Installation of the Market Data Sample App, an app that reads from an external source (market data from dxfeed ) and adds data on data platform.
Infrastructure: Tanzu Kubernetes Grid (TKG) Guest Cluster on VMware Cloud Foundation
You will need to have VMware Cloud Foundation configured following the practices as stated on VMware Validated Design - VVD.
We added some pre-flight check steps on your VMware Cloud Foundation cluster before installing the Data Platform. See README on $BASE_DIR/kubernetes/infrastructure/vvd.
But at very high level, you will need to apply a TanzuKubernetesCluster manifest file and generate a kubeconfig.
cd $BASE_DIR/kubernetes/infrastructure/vvd
# login on the supervisor cluster before running the following
kubectl apply -f dp-k8s-cluster-tkc.yaml
Sample Data Platform Components
Once the Tanzu Kubernetes Cluster is running with at least 8 workers in Ready status, you can install the data platform components with install script - datacomponents/setup-dp-components.sh. Please check the README on $BASEDIR/kubernetes/datacomponents for more details on the validation of the data platform setup.
Dependencies: Please ensure the following command line utilities are present in your system before you proceed with the installation:
- kubectl
- git
cd $BASE_DIR/kubernetes
cp $BASE_DIR/kubernetes/datacomponents/setup-dp-components.sh .
sh setup-dp-components.sh -c allBy the end of this execution, you will have Bitnami Kafka (default), Spark, Solr and Elastic Search installed.
Market Data Sample App
Once you stood up the sample data platform components, you can install the market data sample app that consume market data from dxfeed.com and publish kafka topics and write on Solr using install script - $BASE_DIR/kubernetes/marketdata-sampleapp/setup-marketdata-sampleapp.sh. You can check the README on $BASE_DIR/kubernetes/marketdata-sampleapp for more details on the validation of the market data sample app.
cd $BASE_DIR/kubernetes
cp $BASE_DIR/kubernetes/marketdata-sampleapp/setup-marketdata-sampleapp.sh .
sh setup-marketdata-sampleapp.sh
Changelog
Version Update 1.1
- Bug fix for storage class for bitnami kafka
Contributors
Similar Flings
No similar flings found. Check these out instead...
Demo Appliance for Tanzu Kubernetes Grid
A Virtual Appliance that pre-bundles all required dependencies to help customers in learning and deploying standalone Tanzu Kubernetes Grid (TKG) clusters running on either VMware Cloud on AWS and/or vSphere 6.7 Update 3 environment for Proof of Concept, Demo and Dev/Test purposes.
ESXi Mac Learning dvFilter
This ESX extension adds functionality to ESX to support MAC-learning on vswitch ports. For most ESX use cases, MAC learning is not required as ESX knows exactly which MAC address will be used by a VM.
VCD-nclient
vcd-nclient aims to be a quick start for using the new Notifications feature in vCloud Director v1.5.
SQL30 - An ORM for SQLITE on ESX
SQL30 is a ZERO weight ORM for SQLITE database written using only native python constructs. This ORM works as is on current ESX version of Python without having any additional dependencies.
Autopology - WYSIWYG Designer for NSX
To ease the network creation process and the workloads’ attachment to such networks we present ‘Autopology’ - a graphical interface aimed to complement NSX and simplify workflows for customers and people stepping into the realm of virtual networking.
Linux Driver for Precision Clock Virtual Device
ptp_vmw is a Linux driver for VMware Precision Clock, a new type of virtual device available in ESXi 7.0 (hardware version 17 on-wards) that provides virtual machines with access to the underlying ESXi host's system clock.