Running Pinot in Kubernetes

Pinot quick start in Kubernetes

1. Prerequisites

This quickstart assumes that you already have a running Kubernetes cluster. Please follow the links below to set up a Kubernetes cluster.

• Enable Kubernetes on Docker-Desktop

• Install Minikube for local setup (make sure to run with enough resources e.g. minikube start --vm=true --cpus=4 --memory=8g --disk-size=50g)

• Setup a Kubernetes Cluster using Amazon Elastic Kubernetes Service (Amazon EKS)

• Setup a Kubernetes Cluster using Google Kubernetes Engine (GKE)

• Setup a Kubernetes Cluster using Azure Kubernetes Service (AKS)

2. Setting up a Pinot cluster in Kubernetes

Before continuing, please make sure that you've downloaded Apache Pinot. The scripts for the setup in this guide can be found in our open source project on GitHub.

The scripts can be found in the Pinot source at ./pinot/kubernetes/helm

Git clone project source

# checkout pinot
git clone https://github.com/apache/pinot.git
cd pinot/kubernetes/helm

2.1 Start Pinot with Helm

Run Helm with Pre-installed Package

Pinot repo has pre-packaged HelmCharts for Pinot and Presto. Helm Repo index file is here.

helm repo add pinot https://raw.githubusercontent.com/apache/pinot/master/kubernetes/helm
kubectl create ns pinot-quickstart
helm install pinot pinot/pinot \
    -n pinot-quickstart \
    --set cluster.name=pinot \
    --set server.replicaCount=2

NOTE: Please specify StorageClass based on your cloud vendor. For Pinot Server, please don't mount blob store like AzureFile/GoogleCloudStorage/S3 as the data serving file system.

Only use Amazon EBS/GCP Persistent Disk/Azure Disk style disks.

• For AWS: "gp2"

• For GCP: "pd-ssd" or "standard"

• For Azure: "AzureDisk"

• For Docker-Desktop: "hostpath"

Run Helm Script within Git Repo

2.2 Check Pinot deployment status

kubectl get all -n pinot-quickstart

3. Load data into Pinot using Kafka

3.1 Bring up a Kafka cluster for real-time data ingestion

For Helm v3.0.0

helm repo add incubator https://charts.helm.sh/incubator
helm install -n pinot-quickstart kafka incubator/kafka --set replicas=1

For Helm v2.12.1

3.2 Check Kafka deployment status

kubectl get all -n pinot-quickstart | grep kafka

Ensure the Kafka deployment is ready before executing the scripts in the following next steps.

pod/kafka-0                                                 1/1     Running     0          2m
pod/kafka-zookeeper-0                                       1/1     Running     0          10m
pod/kafka-zookeeper-1                                       1/1     Running     0          9m
pod/kafka-zookeeper-2                                       1/1     Running     0          8m

3.3 Create Kafka topics

The scripts below will create two Kafka topics for data ingestion:

kubectl -n pinot-quickstart exec kafka-0 -- kafka-topics --zookeeper kafka-zookeeper:2181 --topic flights-realtime --create --partitions 1 --replication-factor 1
kubectl -n pinot-quickstart exec kafka-0 -- kafka-topics --zookeeper kafka-zookeeper:2181 --topic flights-realtime-avro --create --partitions 1 --replication-factor 1

3.4 Load data into Kafka and create Pinot schema/tables

The script below will deploy 3 batch jobs.

• Ingest 19492 JSON messages to Kafka topic flights-realtime at a speed of 1 msg/sec

• Ingest 19492 Avro messages to Kafka topic flights-realtime-avro at a speed of 1 msg/sec

• Upload Pinot schema airlineStats

• Create Pinot table airlineStats to ingest data from JSON encoded Kafka topic flights-realtime

• Create Pinot table airlineStatsAvro to ingest data from Avro encoded Kafka topic flights-realtime-avro

kubectl apply -f pinot/pinot-realtime-quickstart.yml

4. Query using Pinot Data Explorer

4.1 Pinot Data Explorer

Please use the script below to perform local port-forwarding, which will also open Pinot query console in your default web browser.

This script can be found in the Pinot source at ./pinot/kubernetes/helm/pinot

./query-pinot-data.sh

5. Using Superset to query Pinot

5.1 Bring up Superset

Open superset.yaml file and goto the line showing storageClass. And change it based on your cloud vendor. kubectl get sc will get you the storageClass value for your Kubernetes system. E.g.

• For AWS: "gp2"

• For GCP: "pd-ssd" or "standard"

• For Azure: "AzureDisk"

• For Docker-Desktop: "hostpath"

Then run:

kubectl apply -f superset.yaml

Ensure your cluster is up by running:

kubectl get all -n pinot-quickstart | grep superset

5.2 (First time) Set up Admin account

kubectl exec -it pod/superset-0 -n pinot-quickstart -- bash -c 'flask fab create-admin'

5.3 (First time) Init Superset

kubectl exec -it pod/superset-0 -n pinot-quickstart -- bash -c 'superset db upgrade'
kubectl exec -it pod/superset-0 -n pinot-quickstart -- bash -c 'superset init'

5.4 Load Demo data source

kubectl exec -it pod/superset-0 -n pinot-quickstart -- bash -c 'superset import_datasources -p /etc/superset/pinot_example_datasource.yaml'
kubectl exec -it pod/superset-0 -n pinot-quickstart -- bash -c 'superset import_dashboards -p /etc/superset/pinot_example_dashboard.json'

5.5 Access Superset UI

You can run below command to navigate superset in your browser with the previous admin credential.

./open-superset-ui.sh

You can open the imported dashboard by clicking Dashboards banner and then click on AirlineStats.

6. Access Pinot using Trino

6.1 Deploy Trino

You can run the command below to deploy Trino with the Pinot plugin installed.

helm repo add trino https://trinodb.github.io/charts/

The above command adds Trino HelmChart repo. You can then run the below command to see the charts.

helm search repo trino

In order to connect Trino to Pinot, we need to add Pinot catalog, which requires extra configurations. You can run the below command to get all the configurable values.

helm inspect values trino/trino > /tmp/trino-values.yaml

To add Pinot catalog, you can edit the additionalCatalogs section by adding:

additionalCatalogs:
  pinot: |
    connector.name=pinot
    pinot.controller-urls=pinot-controller.pinot-quickstart:9000

Pinot is deployed at namespace pinot-quickstart, so the controller serviceURL is pinot-controller.pinot-quickstart:9000

After modifying the /tmp/trino-values.yaml file, you can deploy Trino with:

kubectl create ns trino-quickstart
helm install my-trino trino/trino --version 0.2.0 -n trino-quickstart --values /tmp/trino-values.yaml

Once you deployed the Trino, You can check Trino deployment status by:

kubectl get pods -n trino-quickstart

6.2 Query Trino using Trino CLI

Once Trino is deployed, you can run the below command to get a runnable Trino CLI.

6.2.1 Download Trino CLI

curl -L https://repo1.maven.org/maven2/io/trino/trino-cli/363/trino-cli-363-executable.jar -o /tmp/trino && chmod +x /tmp/trino

6.2.2 Port forward Trino service to your local if it's not already exposed

echo "Visit http://127.0.0.1:18080 to use your application"
kubectl port-forward service/my-trino 18080:8080 -n trino-quickstart

6.2.3 Use Trino console client to connect to Trino service

/tmp/trino --server localhost:18080 --catalog pinot --schema default

6.2.4 Query Pinot data using Trino CLI

6.3 Sample queries to execute

• List all catalogs

trino:default> show catalogs;

  Catalog
---------
 pinot
 system
 tpcds
 tpch
(4 rows)

Query 20211025_010256_00002_mxcvx, FINISHED, 2 nodes
Splits: 36 total, 36 done (100.00%)
0.70 [0 rows, 0B] [0 rows/s, 0B/s]

• List All tables

trino:default> show tables;

    Table
--------------
 airlinestats
(1 row)

Query 20211025_010326_00003_mxcvx, FINISHED, 3 nodes
Splits: 36 total, 36 done (100.00%)
0.28 [1 rows, 29B] [3 rows/s, 104B/s]

• Show schema

trino:default> DESCRIBE airlinestats;

        Column        |      Type      | Extra | Comment
----------------------+----------------+-------+---------
 flightnum            | integer        |       |
 origin               | varchar        |       |
 quarter              | integer        |       |
 lateaircraftdelay    | integer        |       |
 divactualelapsedtime | integer        |       |
 divwheelsons         | array(integer) |       |
 divwheelsoffs        | array(integer) |       |
......

Query 20211025_010414_00006_mxcvx, FINISHED, 3 nodes
Splits: 36 total, 36 done (100.00%)
0.37 [79 rows, 5.96KB] [212 rows/s, 16KB/s]

• Count total documents

trino:default> select count(*) as cnt from airlinestats limit 10;

 cnt
------
 9746
(1 row)

Query 20211025_015607_00009_mxcvx, FINISHED, 2 nodes
Splits: 17 total, 17 done (100.00%)
0.24 [1 rows, 9B] [4 rows/s, 38B/s]

7. Access Pinot using Presto

7.1 Deploy Presto using Pinot plugin

You can run the command below to deploy a customized Presto with the Pinot plugin installed.

Helm

helm install presto pinot/presto -n pinot-quickstart

K8s Scripts

The above command deploys Presto with default configs. For customizing your deployment, you can run the below command to get all the configurable values.

helm inspect values pinot/presto > /tmp/presto-values.yaml

After modifying the /tmp/presto-values.yaml file, you can deploy Presto with:

helm install presto pinot/presto -n pinot-quickstart --values /tmp/presto-values.yaml

Once you deployed the Presto, You can check Presto deployment status by:

kubectl get pods -n pinot-quickstart

Sample Output of K8s Deployment Status

7.2 Query Presto using Presto CLI

Once Presto is deployed, you can run the below command from here, or just follow steps 6.2.1 to 6.2.3.

./pinot-presto-cli.sh

6.2.1 Download Presto CLI

curl -L https://repo1.maven.org/maven2/com/facebook/presto/presto-cli/0.246/presto-cli-0.246-executable.jar -o /tmp/presto-cli && chmod +x /tmp/presto-cli

6.2.2 Port forward presto-coordinator port 8080 to localhost port 18080

kubectl port-forward service/presto-coordinator 18080:8080 -n pinot-quickstart> /dev/null &

6.2.3 Start Presto CLI with pinot catalog to query it then query it

/tmp/presto-cli --server localhost:18080 --catalog pinot --schema default

6.2.4 Query Pinot data using Presto CLI

7.3 Sample queries to execute

• List all catalogs

presto:default> show catalogs;

 Catalog
---------
 pinot
 system
(2 rows)

Query 20191112_050827_00003_xkm4g, FINISHED, 1 node
Splits: 19 total, 19 done (100.00%)
0:01 [0 rows, 0B] [0 rows/s, 0B/s]

• List All tables

presto:default> show tables;

    Table
--------------
 airlinestats
(1 row)

Query 20191112_050907_00004_xkm4g, FINISHED, 1 node
Splits: 19 total, 19 done (100.00%)
0:01 [1 rows, 29B] [1 rows/s, 41B/s]

• Show schema

presto:default> DESCRIBE pinot.dontcare.airlinestats;

        Column        |  Type   | Extra | Comment
----------------------+---------+-------+---------
 flightnum            | integer |       |
 origin               | varchar |       |
 quarter              | integer |       |
 lateaircraftdelay    | integer |       |
 divactualelapsedtime | integer |       |
......

Query 20191112_051021_00005_xkm4g, FINISHED, 1 node
Splits: 19 total, 19 done (100.00%)
0:02 [80 rows, 6.06KB] [35 rows/s, 2.66KB/s]

• Count total documents

presto:default> select count(*) as cnt from pinot.dontcare.airlinestats limit 10;

 cnt
------
 9745
(1 row)

Query 20191112_051114_00006_xkm4g, FINISHED, 1 node
Splits: 17 total, 17 done (100.00%)
0:00 [1 rows, 8B] [2 rows/s, 19B/s]

8. Deleting the Pinot cluster in Kubernetes

kubectl delete ns pinot-quickstart