1. Install Pinot
Run below command to install Pinot using HelmCharts.
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
You can create an example table by running the command below, if you are under pinot binary directory
bin/pinot-admin.sh BootstrapTable -dir examples/batch/airlineStats
2. Install Trino
Run below command to add Trino Helm repo and fetch default helm config file:
helm repo add trino https://trinodb.github.io/charts/
helm inspect values trino/trino > /tmp/trino.yaml
Edit /tmp/trino.yaml file to add your pinot catalog:
Here, since pinot is deployed at namespace pinot-quickstart, so the controller url is pinot-controller.pinot-quickstart:9000
additionalCatalogs:
pinot: |
connector.name=pinot
pinot.controller-urls=pinot-controller.pinot-quickstart:9000
Install Trino HelmCharts with config file:
helm create ns trino-quickstart
helm install my-trino trino/trino --version 0.2.0 --values /tmp/trino.yaml -n trino-quickstart
3. Connecting to Trino
Download Trino Client.
curl -L https://repo1.maven.org/maven2/io/trino/trino-cli/363/trino-cli-363-executable.jar -o /tmp/trino && chmod +x /tmp/trino
Port forward Trino service to your local if it's not already exposed.
export POD_NAME=$(kubectl get pods --namespace trino-quickstart -l "app=trino,release=my-trino,component=coordinator" -o jsonpath="{.items[0].metadata.name}")
echo "Visit http://127.0.0.1:8080 to use your application"
kubectl port-forward $POD_NAME 8080:8080 --namespace trino-quickstart
Use Trino console client to connect to Trino service
/tmp/trino --server localhost:8080 --catalog pinot --schema default
Then write your own queries:
trino:default> show tables;
Table
---------------
airlinestats
baseballstats
(2 rows)
Query 20211022_214646_00003_gmppt, FINISHED, 3 nodes
Splits: 36 total, 36 done (100.00%)
0.44 [2 rows, 59B] [4 rows/s, 133B/s]
trino:default> select count(*) as flights_from_ca_to_ny from airlinestats where originstate='CA' and deststate='NY';
flights_from_ca_to_ny
-----------------------
68
(1 row)
Query 20211022_215024_00007_gmppt, FINISHED, 1 node
Splits: 17 total, 17 done (100.00%)
2.14 [1 rows, 9B] [0 rows/s, 4B/s]
trino:default> select * from airlinestats limit 1;
flightnum | origin | quarter | lateaircraftdelay | divactualelapsedtime | divwheelsons | divwheelsoffs | airtime | arrdel15 | divtotalgtimes | deptimeblk | destcitymarketid | divairportseqids | dayssinceepoch | deptime | month | crselapsedtime | deststatename | carrier | destairportid | distance | arrtimeblk | divarrdelay | securitydelay | longestaddgtime | originwac | wheelsoff | destairportseqid | uniquecarrier | divreacheddest | diverted | actualelapsedtime | airlineid
-----------+--------+---------+-------------------+----------------------+--------------+---------------+-------------+-------------+----------------+------------+------------------+--------------------+----------------+---------+-------+----------------+---------------+---------+---------------+----------+------------+-------------+---------------+-----------------+-----------+-----------+------------------+---------------+----------------+----------+-------------------+-----------
24 | SFO | 1 | -2147483648 | 1968 | [1339, 2310] | [1908, 1758] | -2147483648 | -2147483648 | [121, 50] | 0700-0759 | 31703 | [1319801, 1072102] | 16075 | 723 | 1 | 335 | New York | AA | 12478 | 2586 | 1500-1559 | 1651 | -2147483648 | -2147483648 | 91 | 737 | 1247802 | AA | 1 | 1 | -2147483648 | 19805
(1 row)
Query 20211022_215050_00008_gmppt, FINISHED, 2 nodes
Splits: 48 total, 19 done (39.58%)
1.90 [2 rows, 1.65KB] [1 rows/s, 887B/s]
Meanwhile you can access Trino Cluster UI to see query stats.
